FORECASTING SOFTWARE SURVEY: New tools, capabilities, trends
June 2016
Volume 43 • Number 3 ormstoday.informs.org
Ag analytics Optimizing crop management
Primer on p-values
Reforming America’s
What you need to know
Defining analytics
defense procurement
A conceptual framework
Sports analytics Taxonomy, V1.0
Military spending needs systematic review, but conflicts of interest present major challenges
ORION delivers Edelman for UPS
Contents June 2016 | Volume 43, No. 3 | ormstoday.informs.org
22 On the Cover The Cost of Freedom An Air Force F-16 receives fuel from a KC-135R over Iraq. Can serious analysis reform U.S. military spending and overcome conflicts of interest?
F e at ure s 18
22
Source: Air Force photo by Staff Sgt. Douglas Ellis
Reforming America’s defense procurement Hawks and doves agree U.S. military spending needs a serious, systematic review, but widespread conflicts of interest present a major challenge. By Douglas A. Samuelson
Edelman Award: ‘ORION’ delivers for UPS Revolutionary routing system boosts driver efficiency, cost savings, customer service and the environment. By Peter Horner
26
Ag analytics: Optimizing crop management
30
P ∈ OR (P-values in operations research)
“Smart” application of fertilizer illustrates payoff in using analytical tools to enhance crop yields and improve the environment. By Joseph Byrum
2 | ORMS Today
You don’t need a license to download R, but you should have a good understanding of p-values. By Scott Nestler and Harrison Schramm
|
June 2016
de partm e nt s
6 Inside Story
8 President’s Desk
10 INFORMS in the News
12 Issues in Education
14 PuzzlOR
16 Newsmakers
62 Industry News
63 Classifieds
64 ORacle
12 ormstoday.informs.org
XLMINER®: Data Mining Everywhere Predictive Analytics in Excel, Your Browser, Your Own App
XLMiner® in Excel – part of Analytic Solver® Platform – is the most popular desktop tool for business analysts who want to apply data mining and predictive analytics. And now it’s available on the Web, and in SDK (Software Development Kit) form for your own applications.
Forecasting, Data Mining, Text Mining in Excel. XLMiner does it all: Text processing, latent semantic analysis, feature selection, principal components and clustering; exponential smoothing and ARIMA for forecasting; multiple regression, k-nearest neighbors, and ensembles of regression trees and neural networks for prediction; discriminant analysis, logistic regression, naïve Bayes, k-nearest neighbors, and ensembles of classification trees and neural nets for classification; and association rules for affinity analysis.
have in Excel, and generate the same reports, displayed in your browser or downloaded for local use.
XLMiner SDK: Predictive Analytics in Your App. Access all of XLMiner’s parallelized forecasting, data mining, and text mining power in your own application written in C++, C#, Java, R or Python. Use a powerful object API to create and manipulate DataFrames, and combine data wrangling, training a model, and scoring new data in a single operation “pipeline”.
Find Out More, Start Your Free Trial Now. Visit www.solver.com to learn more, register and download Analytic Solver Platform or XLMiner SDK. And visit www.xlminer.com to learn more and register for a free trial subscription – or email or call us today.
XLMiner.com: Data Mining in Your Web Browser. Use a PC, Mac, or tablet and a browser to access all the forecasting, data mining, and text mining power of XLMiner in the cloud. Upload files or access datasets already online. Use the same Ribbon and dialogs you
The Leader in Analytics for Spreadsheets and the Web Tel 775 831 0300 • info@solver.com • www.solver.com
June 2016 | Volume 43, No. 3 | ormstoday.informs.org
44
INFORMS Board of Directors
President Edward H. Kaplan, Yale University President-Elect Brian Denton, University of Michigan
Past President L. Robin Keller, University of California, Irvine
Secretary Pinar Keskinocak, Georgia Tech
Treasurer Sheldon N. Jacobson, University of Illinois Vice President-Meetings Ronald G. Askin, Arizona State University Vice President-Publications Jonathan F. Bard, University of Texas-Austin
F e at ure s 34
Vice President- Esma Gel, Arizona State University Sections and Societies
Vice President- Marco Lüebbecke, Information Technology RWTH Aachen University
Defining analytics: a conceptual framework Analytics’ rapid emergence a decade ago created a great deal of corporate interest, as well as confusion regarding its meaning. By Robert Rose
Vice President- Jonathan Owen, CAP, General Motors Practice Activities Vice President- Grace Lin, International Activities Institute for Information Industry
Vice President-Membership Susan E. Martonosi, Professional Recognition Harvey Mudd College Vice President-Education Jill Hardin Wilson, Northwestern University Vice President-Marketing, Laura Albert McLay, Communications and Outreach University of Wisconsin-Madison Vice President-Chapters/Fora Michael Johnson, University of Massachusetts-Boston
40
Sports analytics taxonomy, V1.0 Classification techniques provide an important first step in the serious study of the fast-growing field of sports analytics. By Gary Cokins, Walt DeGrange, Stephen Chambal and Russell Walker
Editors of Other INFORMS Publications Decision Analysis Rakesh K. Sarin, University of California, Los Angeles
I NFORMS Journal on Computing David Woodruff, University of California, Davis
Co m puting 44
Editor’s Cut Anne G. Robinson, Verizon Wireless
Information Systems Research Ritu Agarwal, University of Maryland
INFORMS Online Kevin Geraghty, 360i INFORMS Transactions Jeroen Belien, KU Leuven on Education
Interfaces Srinivas Bollapragada, General Electric Global Research Center
Software survey: Forecasting 2016 New tools, new capabilities and new trends: Survey of 26 software packages from 19 vendors. By Chris Fry and Vijay Mehrotra
n ews
Management Science Teck-Hua Ho, National University of Singapore Office of the Deputy President (Research and Technology)
Manufacturing & Service Christopher S. Tang, Operations Management University of California, Los Angeles
Marketing Science K. Sudhir, Yale University
Mathematics of Operations J. G. “Jim” Dai, Cornell University Research
Operations Research Stefanos Zenios, Stanford University
Organization Science Zur Shapira, New York University
Service Science Paul P. Maglio, University of California, Merced Strategy Science Daniel A. Levinthal, Wharton School, University of Pennsylvania Transportation Science Martin Savelsbergh, Georgia Institute of Technology
55 Annual Meeting set for Nashville 57 Carnegie Mellon wins UPS Prize
57 MIT team earns IAAA honors
58 Roundtable spring roundup
4 | ORMS Today
|
June 2016
58 DMDA Workshop
59 Call for nominations
60 People
61 Winter Simulation Conference
61 Meetings
Tutorials in Operations J. Cole Smith, University of Florida Research
INFORMS Office • Phone: 1-800-4INFORMS
Executive Director Melissa Moore
Headquarters
INFORMS (Maryland) 5521 Research Park Dr., Suite 200 Catonsville, MD 21228 USA Tel.: 443.757.3500 Fax: 443.757.3515 E-mail: informs@informs.org
ormstoday.informs.org
Su pp ac Pow orts he e T Sp r B abl ar I a ea k n u ExcelBig Dd , at a
Ap
ANALYTIC SOLVER PLATFORM ®
From Solver to Full-Power Business Analytics in
Solve Models in Desktop Excel or Excel Online.
Plus Forecasting, Data Mining, Text Mining.
From the developers of the Excel Solver, Analytic Solver Platform makes the world’s best optimization software accessible in Excel. Solve your existing models faster, scale up to large size, and solve new kinds of problems. Easily publish models from Excel to share on the Web.
Analytic Solver Platform samples data from Excel, PowerPivot, and SQL databases for forecasting, data mining and text mining, from time series methods to classification and regression trees and neural networks. And you can use visual data exploration, cluster analysis and mining on your Monte Carlo simulation results.
Conventional and Stochastic Optimization. Fast linear, quadratic and mixed-integer programming is just the starting point in Analytic Solver Platform. Conic, nonlinear, non-smooth and global optimization are just the next step. Easily incorporate uncertainty and solve with simulation optimization, stochastic programming, and robust optimization – all at your fingertips.
Find Out More, Download Your Free Trial Now. Analytic Solver Platform comes with Wizards, Help, User Guides, 90 examples, and unique Active Support that brings live assistance to you right inside Microsoft Excel. Visit www.solver.com to learn more, register and download a free trial – or email or call us today.
Fast Monte Carlo Simulation and Decision Trees. Analytic Solver Platform is also a full-power tool for Monte Carlo simulation and decision analysis, with 50 distributions, 40 statistics, Six Sigma metrics and risk measures, and a wide array of charts and graphs.
The Leader in Analytics for Spreadsheets and the Web Tel 775 831 0300 • info@solver.com • www.solver.com
Inside Story
Peter Horner, editor
peter.horner@mail.informs.org
OR/MS Today Advertising and Editorial Office
Queue tips from Dr. Queue
Send all advertising submissions for OR/MS Today to: Lionheart Publishing Inc. 506 Roswell Street, Suite 220, Marietta, GA 30060 USA Tel.: 888.303.5639 • Fax: 770.432.6969
President
In the weeks leading up to Memorial Day weekend, hardly a day went by without the mainstream media in U.S. cities with major international airports sending a reporter, film crew or photographer to the city’s airport to document the incredibly long lines of passengers trying to get through TSA checkpoints. The reporters inevitably reported that if you think this is bad, just wait until the Memorial Day weekend. Cue Dr. Queue. Dick Larson, aka “Dr. Queue,” is a professor at MIT and a past president of INFORMS whose career has focused on operations research as applied to services industries. Larson’s many years of research on queues has made him the go-to guy for reporters looking for an expert on the science, psychology and pain of long lines, whether it’s waiting for an elevator, a Disney World ride, a grocery store check-out or an airport check-in. Right on cue, reporters from several major media outlets began checking in with Dr. Queue for his thoughts on the airport issues in May. I live in the Atlanta area, and ATL suffered some of the longest queues in the country, a situation aggravated by the closing of a key security checkpoint for renovations and restructuring in May. Like other reporters, I planned to ask Dr. Queue whether I really needed to waste three hours of my life hanging around the airport to ensure that I made my next flight. But other reporters beat me to the punch. Turns out Dr. Queue had already been interviewed by several members of the media on this very subject. I n a M ay 1 7 p o s t i n g , W B E Z in Chicago introduced a lengthy
inter view with Larson as follows: “No doubt plenty of laymen in lines at O’Hare and Midway are whittling away the hours pointing out ways their lines could be moving faster. So, do they have a point? We ask an expert on ‘queuing theory’ about the best ways to organize and manage lines in public places like airports. Dick Larson is a professor of engineering systems at MIT, but we like his superhero name much better: Dr. Queue.” In a May 18 posting on the Science of Us entitled “The Strange Science of Why Airport Security Lines Spiral Out of Control,” writer Drake Baer notes: “A lot of this, Larson says, has to do with the profoundly human quality of variability. If queues were mechanical – like in a well-run factory, where the time of arrival and the time of service for each transaction were highly predictable – then a server could be super busy and queues still wouldn’t form.” In a May 20 online ar ticle for NBC News, reporter Harriet Baskas writes that [Larson] “says the circus enter tainer s, therapy ponies, live music and free snacks some airports are offering to those waiting in long checkpoints lines could backfire. It works for Disney in the amusement parks, said Larson. But passengers who miss flights due to long checkpoint lines may end up being more furious ‘because they’ll feel like they were being distracted from what’s really important – getting on the plane.’ ” As for ATL, airport officials brought in a bunch of additional TSA folks, reopened the checkpoint, and waiting lines were reduced to 15 or 20 minutes over the dreaded Memorial weekend. Now why didn’t I think of that? ORMS
— Peter Horner, editor peter.horner@mail.informs.org
6 | ORMS Today
|
June 2016
John Llewellyn, ext. 209 john.llewellyn@mail.informs.org
Editor Peter R. Horner peter.horner@mail.informs.org Tel.: 770.587.3172
Assistant Editor Donna Brooks
Contributing writers/editors Douglas Samuelson, Matt Drake, John Toczek
Art Director Alan Brubaker, ext. 218 alan.brubaker@mail.informs.org
Online Projects Manager Patton McGinley, ext. 214 patton.mcginley@mail.informs.org
Assistant Online Projects Manager Leslie Proctor, ext. 228 leslie.proctor@mail.informs.org
Advertising Sales Managers Sharon Baker sharon.baker@mail.informs.org Tel.: 813-852-9942 Aileen Kronke, ext. 212 aileen@lionhrtpub.com
Reprints Kelly Millwood, ext. 215 kelly.millwood@mail.informs.org
OR/MS Today Committee James Cochran, chairman
INFORMS Online http://www.informs.org
Lionheart Publishing Online http://www.orms-today.org OR/MS Today (ISSN 1085-1038) is published bimonthly by the Institute for Operations Research and the Management Sciences (INFORMS). Canada Post International Publications Mail (Canadian Distribution) Sales Agreement No. 1220047. Deadlines for contributions: Manuscripts and news items should arrive no later than six weeks prior to the first day of the month of publication. Address correspondence to: Editor, OR/MS Today, 506 Roswell Street, Suite 220, Marietta, GA 30060. The opinions expressed in OR/MS Today are those of the authors, and do not necessarily reflect the opinions of INFORMS, its officers, Lionheart Publishing Inc. or the editorial staff of OR/MS Today. Membership subscriptions for OR/MS Today are included in annual dues. INFORMS offers non-member subscriptions to institutions, the rate is $62 USA, $79 Canada & Mexico and $85 all other countries. Single copies can be purchased for $10.50 plus postage. Periodicals postage paid at Catonsville, MD, and additional mailing offices. Printed in the United States of America. POSTMASTER: Send address changes to OR/MS Today, INFORMS-Maryland Office, 5521 Research Park Dr., Suite 200, Catonsville, MD 21228. OR/MS Today copyright ©2016 by the Institute for Operations Research and the Management Sciences. All rights reserved.
ormstoday.informs.org
Your Analytics App – Everywhere
Use Solver, Risk Solver, XLMiner in Excel Online, Google Sheets Or Turn YOUR Excel Model into a Web or Mobile App in Seconds
The easiest way to build an analytic model – in Excel – is now the easiest way to deploy your analytic application to Web browsers and mobile devices – thanks to the magic of Frontline Solvers® and our RASON® server.
Use our Analytics Tools in your Web Browser. Solve linear, integer and nonlinear optimization models with Frontline’s free Solver, and run Monte Carlo simulation models with our free Risk Solver® tool, in Excel Online and Google Sheets. Use our free XLMiner® Analysis ToolPak tool for statistical analysis, matching the familiar Analysis ToolPak in desktop Excel.
Build Your Own Apps with RASON Software. RASON – RESTful Analytic Solver® Object Notation – is a new modeling language for optimization and simulation that’s embedded in JSON (JavaScript Object Notation). With support for linear, nonlinear and stochastic optimization, array and vector-matrix operations, and dimensional tables linked to external databases, the RASON language gives you all the power you need.
Your Excel Model Can Be a Web/Mobile App. The magic begins in Excel with Frontline Solvers V2016: Our Create App button converts your Excel optimization or simulation model to a RASON model, embedded in a Web page, that accesses our cloud servers via a simple REST API. You’re ready to run analytics in a browser or mobile device! Or if you prefer, run your RASON model on your desktop or server, with our Solver SDK®. Either way, you’re light-years ahead of other software tools.
Find Out More, Sign Up for a Free Trial Now. Visit www.solver.com/apps to learn more, and visit rason.com to sign up for a free trial of RASON and our REST API. Or email or call us today.
The Leader in Analytics for Spreadsheets and the Web Tel 775 831 0300 • info@solver.com • www.solver.com
President’s Desk
‘Me m
b e r-in- C h ie f Me m o’
Ed Kaplan
INFORMS President president@informs.org
Sexy operations research Yale’s School of Management is sending about 26 students to jobs or internships at Amazon this summer, so interest was already high when Ed McGavin and Murali Rajamani strolled into a classroom packed with MBAs for a noontime presentation. Representing Amazon Logistics, McGavin (Ph.D. in operations research from Purdue’s Krannert school) is their process engineering leader, while Rajamani (Ph.D. in chemical engineering from Wisconsin and an MBA from Chicago Booth) leads their operations tech implementation team. As Ed and Murali eased into their talk on the topology and processes underlying Amazon’s fulfillment network, a glance around the room captured student reactions not often seen in such sessions. Said students were nearly salivating from the edge of their seats as Ed and Murali connected facility layout, information systems and their variant of the traveling salesman problem.The audience was not only finding the material interesting; the expression on their faces conveyed an unmistakable sentiment:This stuff is hot! This reaction is not a one-off. Those same awestruck looks could be seen during the Edelman Prize Gala finalist presentations at the highly successful INFORMS Conference on Business Analytics and Operations Research that took place in Orlando, Fla., this past April, especially when the winning team from United Parcel Service (UPS) showcased their On-Road Integrated Optimization and Navigation (ORION) system.As impressive as its technical accomplishments is the depth to which this project permeates the entire workforce at UPS.This fact hit home when Carl (your member-inchief’s UPS delivery person), upon learning of his company’s big Edelman win, pulled out his PDA to proudly display the rest of his ORION-generated schedule for the afternoon. He also remarked how great it was that the system helped him avoid making all those left turns. Carl loves ORION too. 8 | ORMS Today
|
June 2016
David Abney, chairman and CEO of UPS, has certainly been sold on the value of operations research within his own organization. In conversations with Abney and other top UPS brass, ORION team leader and INFORMS President’s Award winner Jack Levis made a rather important point: While fancy modeling gives very smart O.R. types the opportunity to showcase their skill, no matter how cool such work is, it is unlikely to be adopted by private sector organizations unless it solves their business problems. By extension, getting O.R. embedded better in public and not-for-profit organizations similarly requires the resulting analysis to be helpful in some important way. INFORMS recognizes the importance of having organizations adopt analytics and operations research. After all, operations research is all about decision-making, and most important, decision-makers reside in organizations! Indeed, one of the four goals in the INFORMS Strategic Plan states that organizations will identify operations research and analytics as core components of success and institutionalize their input as part of their decision-making processes. How is INFORMS tackling this goal? There are several points of contact between INFORMS and other organizations, most of which fall under the purview of the vice president of Practice Activities, a position presently (and ably!) filled by Jonathan Owen, director of operations research at General Motors (which, by the way, received the 2016 INFORMS Prize for effective integration of advanced analytics and OR/MS in an organization). Part of the PracticeVP’s charge includes evaluation of the impact of INFORMS programs and activities on practitioners and organizations, to quote our Talmudic Policies and Procedures Manual (Section 12 J). The VP-Practice is the liaison between the INFORMS Board and the INFORMS Roundtable, a consortium of 50+ organizations, each represented by a seasoned OR/MS professional. Currently chaired by
Bill Browning of Applied Mathematics, Inc., the Roundtable provides direct access to the wisdom and experience of top OR/MS leaders who are influential in their home organizations’ employment of O.R. ideas. Indeed, part of the Roundtable’s stated mission is to “improve member organizations through superior OR/MS performance [1]. Also reporting to the VP-Practice is the new Committee on Engagement with Organizations. Chaired by Zahir Balaporia of FICO, and quoting literally from the committee’s charge, this committee was formed to develop and advocate activities and services for, and increase engagement with, organizations. Said committee is externally focused with the goal of increasing awareness of INFORMS and its value to organizations. One more committee reporting to our VP-Practice is the Analytics Maturity Model (AMM) Committee. Again with reference to the committee’s charge, the AMM helps organizations conduct a selfassessment of how they use analytics. The idea is to help organizations (really analytics missionaries within organizations) design a plan that can help them improve their use of analytics.The AMM Committee’s job is to create the AMM, update it as needed, and bring it to the attention of those who might benefit from it. It is fantastic that students and some professionals have discovered how exciting O.R. can be. It would be even better if senior decision-makers in leading organizations, whether private, public or not-for-profit, recognize that the benefits of employing O.R. and analytics in their establishments are much hotter than the methods themselves. If our approach to doing stuff with organizations succeeds, INFORMS can make operations research and analytics irresistible! ORMS REFERENCE 1. https://www.informs.org/Community/Roundtable/ Vision-Missions-and-Goals
ormstoday.informs.org
Is your solver provider
getting in the way of your success?
At Gurobi, we make it easier to achieve your optimization goals. Faster solutions with superior algorithms and powerful tuning Maximum productivity with flexible interfaces and modeling language support Great technical support from easy-to-reach optimization experts No surprises with our transparent pricing and flexible licensing
Gurobi Optimizer 6.5 State-of-the-Art Mathematical Programming Solver Experience state-of-the-art optimization algorithms backed by industry-leading customer support.
www.gurobi.com +1 713-871-9341 info@gurobi.com
Contact us today to request a FREE full-featured, unrestricted evaluation license.
INFORMS in the News
Compiled by Olivia Schmitz, marketing coordinator at INFORMS.
Library of things, ORION and airlines INFORMS members, initiatives and journals continue to make news on a wide range of topics in a variety of media forums. Following are recent excerpts of “INFORMS in the News”: CAP in the U.K. Michael Mortenson, a Ph.D. student at Loughborogh University researching the development of analytics education in U.K. universities and a CAP® certified professional, recommends the CAP exam “both to analytics/O.R. professionals seeking to ‘prove’ their practical expertise, and to employers looking for recruits who can genuinely hit the ground running.” - Impact, April 2016
Libraries are reducing clutter, increasing environmental sustainability You’ve heard of the Internet of Things; now introducing the library of things – a new program at local libraries where people can check out useful items like sewing machines, printers, tools and more.This is an effort to reduce waste and build a “sharing economy” – the business model similar to that of AirBnB and Uber. [INFORMS member] Saif Benjaafar, professor of engineering and director of the sharing economy initiative at the University of Minnesota, says that, “Economy-wide, there is significant waste associated with these cheaper items, as they tend to be poorly maintained and frequently replaced. The concept of Library of Things has the potential of significantly reducing such waste.” - Huffington Post, April 29
ROI for UPS’ ORION project decreases costs with predictive analytics Tom Davenport provides his “marginal” description of predictive analytics and the ORION (On-Road Integrated Optimization and Navigation) Project, created by the parcel delivery company UPS. “It’s basically prescriptive analytics 10 | ORMS Today
|
June 2016
for UPS’ 55,000 drivers in the United States (the international rollouts will come soon). Instead of following the same route every day, ORION analyzes the packages to be delivered that day and determines an optimal routing for the “package cars.”The drivers are told where to go next by their handheld computers.” Davenport also provides an answer to: What does ORION do? “This project won the 2016 Franz Edelman Award which recognizes excellence in operations research and analytics in the public and private sectors that change the involved organizations.” - Tom Davenport in Data Informed, April 19
What can predictive analytics do for airlines? PROS, the revenue and profits realization company, showcased its new application of operations research methodology for the airline industry at the AGIFORS Revenue Management Conference in Frankfurt, Germany, on May 18-20. According to the Business Wire article published on May 11, the project – “A Capacity-Sharing Model with Movable Curtain” – was developed from research gathered from PROS airline customers. “It is a revenue management model in which an aircraft’s seating capacity is shared between business and economy compartments. More specifically, the respective compartment capacities can be adjusted by means of curtain installed shortly before departure, depending on demand realization, and economy passengers can be dynamically considered for accommodation in the business compartment, which leads to better utilization of the aircraft’s capacity and improved profitability.” - Business Wire, May 11
Virginia Tech honors industrial & systems engineers The Grado Department of Industrial & Systems Engineering (ISE) and Virginia Tech honored three distinguished alumni on April 25. Among the three winners is INFORMS member Dr. Janis P.Terpenny. Dr.Terpenny is a pillar as an academic in the analytics and operations research industry. - Roanoke Times, May 12
ASU honors leader in engineering education Arizona State University, a leader in eng ineer ing education, honor s [INFORMS member] Ron Askin as a dedicated, driving force behind its rise and recognition. Askin became the director of the School of Computing, Informatics and Decision Systems Engineering (CIDSE) in 2009. Since then, he has been successful in strengthening the discipline and increasing the size of the student body. - ASU News, May 12
Syngenta and INFORMS award Crop Challenge winners Analytics and operations research can have a huge impact on science techniques in agriculture. Just ask the 2016 Syngenta Crop Challenge winners: Stanford University.Their entry,“Hierarchy modeling of soybean variety yield and decision making for future planting plan,” was awarded the $5,000 Syngenta Crop Challenge prize at the INFORMS Conference on Business Analytics and Operations Research. “Operations research and advanced analytics can contribute to variety development and evaluation, reducing costs and improved efficiency,” said Xiaocheng Li, a member of the Stanford team. - Farm Futures, April 26
Twitter identifies if your brand image is green or healthy What do Toyota, Aveda and Clif Bar have in common? An article in the INFORMS journal Marketing Science finds that Twitter fans of these brands are all more likely to follow accounts that tweet about the environment.This in turn creates a greener image than other brands in their sectors. - Management Science, April 18 ORMS ormstoday.informs.org
Most enterprises rely on analytic experts to solve their toughest challenges. But to get ahead of the competition, you need to put powerful optimization capabilities in the hands of your business users — so they can deliver value with no friction or delays.
Take FICO Optimization Modeler for a test drive now, and transform your most complex business challenges into smarter, faster, more profitable customer decisions.
www.fico.com/OMtrial
www.fico.com/optimization FICOAmericas@fico.com Š 2016 Fair Isaac Corporation. All rights reserved.
Issues in Education
Wendy Swenson Roth
Teaching modeling to business students “Modeling is one of the fundamental ways in which human beings understand the world.” [1] – Stephen G. Powell
12 | ORMS Today
|
June 2016
the value in these skills, increasing the pool of interested. Each of the recent times I have taught this class the enrollments have been near or at the maximum allowed. This, I believe, is due to the popularity of business analytics, which has led to increased student i n t e re s t i n bu s i n e s s modeling classes. The result is a broader variety of student backgrounds; not necessarily self-selected based on strong mathematical backgrounds as compared to the student skills level five to 10 years ago when I first taught business modeling. The majority of undergraduate students taking my class are managerial sciences majors who take broad ranges of courses, including such topics as human resources and operations. The graduate students are approximately 50 percent MBAs.The rest are from a variety of areas including masters in managerial sciences and actuarial sciences. In addition to the benefit to students of learning optimization, this increased interest in what can be classified as a math-based, critical thinking class is encouraging for other reasons. According to the 2013 Skills Outlook from the Organization for Economic Cooperation and Development [3]: “Proficiency in literacy, numeracy and problem solving in technology-rich environments is positively and independently associated Image © rawpixel | 123rf.com
In the spring of 2015, for the first time in a number of years, my teaching schedule included two business modeling classes that focused on optimization models. My previous experience was at the graduate level. Now I would be teaching both graduate and undergraduate students. The tools used today to solve optimization programs are very different from the graph paper I used when first learning to solve linear programming models as part of an undergraduate engineering course. With the availability of user-friendly software, what was once a skill taught by engineering departments is now part of the business school curriculum. Since my introduction to optimization, there has been a significant shift in the environment for students. It began in the 1990s with the introduction of readily available software packages to solve these models. Powell describes this change in a 1997 column: “Dan Fylstra (designer of Solver in Excel) points out, more copies of a spreadsheet package with a built-in linear and nonlinear solver are sold every month than there are management scientists in the world. My MBA students routinely set up and solve optimization and simulation problems that I could not have solved (in less than a week) when graduating with a Ph.D. in the early 1980s” [2]. These readily available software packages have changed the skills required for students. In another shift, analytics has drawn more attention to
with the probability of participating in the labor market and being employed, and with higher wages” (page 24). A common definition of problem-solving is: “The process of working through details of a problem to reach a solution. Problem solving may include mathematical or systematic operations and can be a gauge of an individual’s critical thinking skills” [4]. Classes such as business modeling provide excellent opportunities for students to learn these skills – skills that are
desired by hiring managers, but are often not found in students. In a study by U.S. Harris in 2013, only 41 percent of hiring managers not looking for STEM majors say recent graduates are completely or very prepared to solve problems through experimentation. Based on the above information, the way I teach business modeling has evolved to better match my students’ skills and focus on problem-solving not just tool usage. In addition to teaching optimization, other important issues that need to be addressed include: 1. Vary the math skills of students and provide additional explanation and support for students with weaker quantitative skills. 2. Teach students to develop good spreadsheet skills.This is stressed in lecture and with explanations of the pitfalls of poorly constructed spreadsheets, including examples of good and bad spreadsheets. 3. Teach how to de-bug a model, including sanity-checking your answer. ormstoday.informs.org
I have been working on developing some games for students to “play” to reinforce the concepts and give them feedback during the process. Addressing these issues has caused me to rethink how I conduct my class. One action that has helped address some of the above issues is asking students to bring their laptops, thereby allowing us to create the models together.This requires students to get their hands dirty and see if they really understand what we are covering, allowing us to resolve minor issues that might turn in to major frustrations if they wait to resolve them at home. Another way that I have tried to address some of the above issues came about in response to students’ requests for more example problems. After in-
troducing topics, class time is devoted to working out problems; unfortunately, there isn’t time to increase the examples in class. Just giving more textbook problems doesn’t seem like a good solution. I have been working on developing some games for students to “play” to reinforce the concepts and give them feedback during the process instead of just at the end of the game/model when the model is solved. Learning to perform optimization is a great skill. Equally important is the critical thinking and problem-solving skills that can be learned in such a course. My goal has been to provide an environment to help students who may view their math skills as not strong enough to succeed in developing the skill of optimization, and therefore
REFERENCES 1. Powell, S. G., 1995, “The Teacher’s Forum: Teaching the Art of Modeling to MBA Students,” Interfaces, May-June, pp. 88-94. 2. Powell, S. G., 1997, “The Teacher’s Forum: From Intelligent Consumer to Active Modeler, Two MBA Success Stories,” Interfaces, May-June, pp. 88-98. 3. OECD, 2013, “OECD Skills Outlook 2013: First Results from the Survey of Adult Skills,” http:// www.oecd.org/site/piaac/publications.htm. 4. Business Dictionary, 2016, “problem solving,” http://www.businessdictionary.com/definition/ problem-solving.html. 5. Powell, S. G., 1995, “The Teacher’s Forum: Six Key Modeling Heuristics,” Interfaces, July-August, pp. 114-125. 6. Powell, S. G., “Teaching Modeling in Management Science,” INFORMS Transactions on Education, Vol. 1, No. 2, pp. 62-67.
increase their math and critical thinking skills; skills that will help them succeed in their future careers. ORMS Wendy Swenson Roth (wroth@gsu.edu) is a clinical assistant professor of managerial sciences at Georgia State University.
Dynamic Ideas llc New TiTles
AdditionAl titles By dynAmic ideAs Cover_R1_4761.qxd 64446_Hard_cover_45xx.qxd
8/16/05
8:39 PM
Page 1
00_COVER_4770
4/11/07
1:59 PM
8:56 AM
Page 1
Abhijit A. Pol and Ravindra K. Ahuja
Each topic covered is illustrated through examples and hands-on tutorials. Each chapter contains several hands-on exercises for additional practice. This book is ideally suited as a textbook but can also be used as a supplementary reference book or a self-study manual. The book Web site, www.dssbooks.com, contains supplementary material for students and instructors including additional case studies. AUTHORS: Abhijit A. Pol is a researcher in the Department of Computer and Information Science and Engineering at the University of Florida, Gainesville. His research focus is in the area of databases with special interests in approximate query processing and physical database design. Ravindra K. Ahuja is a professor in the Department of Industrial and Systems Engineering at the University of Florida, Gainesville. He is also the President of Innovative Scheduling, Inc., which specializes in building decision support systems for planning and scheduling problems arising in the field of transportation.
DIMITRIS BERTSIMAS ROBERT M.,!F7REUND IA9H5-jbegfg! ISBN 0-9759146-5-0
A b h i j i t A . Po l R av i n d r a K . A h u j a
www.dynamic-ideas.com
DI_ORMS_AD.indd 1
Among its special features, the book: ■
Develops the theory of integer optimization from a new geometric perspective via integral generating sets
■
Emphasizes strong formulations, ways to improve them, integral polyhedra, duality, and relaxations
■
Discusses applications of lattices and algebraic geometry to integer optimization, including Gröbner bases, optimization over polynomials and counting integer points in polyhedra
■
Contains a unified geometric treatment of cutting plane and integral basis methods
■
Covers enumerative and heuristic methods, including local search over exponential neighborhoods and simulated annealing
■
Presents the major methods to construct approximation algorithms: primal-dual, randomized rounding, semidefinite and enumerative methods
■
Provides a unified treatment of mixed integer and robust discrete optimization
■
Includes a large number of examples and exercises developed through extensive classroom use
Optimization over Integers
“For over a quarter of a century, Urban Operations Research has been a primary source for introducing thousands of students to the world of operations research applications. Anyone interested in how a city can improve its critical services will find basic and advanced ideas clearly explained and grounded in practicality. Of special interest is the rare discussion on “Implementation.” Here, the novice student and the practiced researcher will find sound advice that will help ensure that their mathematical models will make a difference. Case in point, ‘Beware of the Vanishing Advocate.’” Saul I. Gass—Professor Emeritus, Robert H. Smith School of Business, University of Maryland, College Park
“Of all the courses I took as an undergraduate and graduate student at M.I.T., Urban Operations Research undoubtedly had the greatest impact on my career and on my way of thinking about the world around me. To this day, over thirty years after taking the course, I often find myself referring to the text for insights and solutions to problems. I would recommend this book to anyone interested in operations research at any level.” Mark S. Daskin—Bette and Neison Harris Professor of Teaching Excellence, Northwestern University
“Having gone through course after course on the theoretical underpinnings of OR, this book opened my eyes as a student to the impact that OR modeling can have on real-world problems. It showed me how rigorous analysis can be applied to address fundamental problems in society. It’s an absolute classic in the field.” Patrick T. Harker—President, University of Delaware
“I still use my totally worn-out copy of the first edition of Urban Operations Research, bought when I was a graduate student at MIT. Dick and Amedeo’s book belongs on the desk of all operations researchers, not only those interested in efficient resource allocation of urban services. It is one of the finest examples of the power of quantitative modeling. The text is a classic and I am delighted to see it re-edited.” Patrick Jaillet—Edmund K. Turner Professor and Department Head, Department of Civil and Environmental Engineering, MIT
“Urban Operations Research introduced me to realistic and practical modeling of very complex problems. What I learned from Amedeo and Dick changed the way I think and my approach to problem solving, setting the direction for my career. I have been using my loose-leaf, pre-publication copy ever since 1978 when I took the course.”
Urban Operations Research
Urban Operations Research
PART V—CASE STUDIES: This part presents several case studies of decision support systems arising in different application settings including Online Book Store, Portfolio Management and Optimization, and Television Advertisement Allocation.
Using Access, VB .NET, and ASP .NET
T
his book provides a new, unified, insightful, comprehensive and modern treatment of integer optimization. It includes classical topics as well as the state of the art, in both theory and practice.
Optimization over Integers
PART III—WINDOWS APPLICATION DEVELOPMENT WITH VB .NET: The topics for this part include the Visual Studio development environment, object-oriented and visual programming, database connectivity using ADO .NET, and Crystal Report .NET.
Using Access, VB .NET, and ASP .NET
PART II—DATABASE DEVELOPMENT WITH MICROSOFT ACCESS: This part describes how to create tables using Microsoft Access, establish relationships between tables, and manipulate tables using Access queries or using Structured Query Language (SQL).
PART IV—WEB APPLICATION DEVELOPMENT WITH ASP .NET: This part illustrates the process of developing Web applications by using HTML and ASP .NET.
Developing Web-Enabled Decision Support Systems
PART I—PRINCIPLES OF GOOD DATABASE DESIGN: This part of the book covers entity-relationship diagrams, creating relational databases, and normalizing databases.
Developing Web-Enabled Decision Support Systems
Larson Odoni
Using Access, VB .NET, and ASP .NET
Developing Web-Enabled Decision Support Systems is a comprehensive book that describes how to build data-driven, Web-enabled decision support systems using a Microsoft Access database, VB .NET, and an ASP .NET framework, and illustrates it using several case studies arising in Operations Research, Industrial Engineering, and Business. The book contains five parts:
Bertsimas Weismantel
Pol Ahuja
Developing Web-Enabled
Decision Support Systems Data, Models, and Decisions
The Fundamentals of Management Science
1/18/07
Page 815
Barry Smith—Senior Vice President and Chief Scientist, Sabre Holdings (widely considered to be an “inventor” of airline revenue management)
“Urban Operations Research is a tremendous resource for improved modeling and decision making in today's dynamic business environment—both an essential text for preparing students and a valuable reference for experienced OR practitioners."
Dimitris Bertsimas is the Boeing Professor of Operations Research at the Massachusetts Institute of Technology and Robert Weismantel is Professor of Mathematics in the University of Magdeburg, Germany.
ISBN 0-9759146-3-4
Dimitris Bertsimas
ISBN number: 0-9759146-2-6
Di
Mitchell Burman—Managing Partner, Analytics Operations Engineering, Inc.
R i c h a r d C. L a r s o n
,!7IA9H5-jbegdc!
A m e d e o R. O d o n i
Robert Weismantel
Phone: 603 882-399 Fax: 603 386-6197
email: orders@altlog.net
June 2016
|
ORMS Today
4/8/16 8:07 PM
| 13
PuzzlOR
John Toczek
puzzlor@gmail.com
Elevators The building I work in has three elevators and 12 floors. Like some elevators, these have displays above them that indicate the current floor that the elevator is on and also the travel direction of the elevator. The travel direction display can show Up, Down or Stopped indicated by an up A. DISPLAY
B. DISPLAY
C. DISPLAY
8? 66 95 8? 11? 76 75 56 25 75 25
56 9? 66 116 55 45 55 85 12? 55 4?
35 35 25 85 85 5? 12? 126 55 12? 66
1 2 3 4 5 6 7 8 9 10 11
arrow, a down arrow or a dash character, respectively.As an example, Elevator A in the image is on Race to the bottom: Which elevator to take? the 2nd floor and is traveling in the up direction. Table 1 displays 10 historic data is most likely to be first to the ground points showing the floor and direction floor? of the elevators from Send your answer to puzzlor@gmail.com FIRST TO times that I have by Aug. 15. The winner, chosen randomly GROUND FLOOR used the elevator in from correct answers, will receive a $25 B the past. The table Amazon Gift Card. Past questions and answers A B also shows which can be found at puzzlor.com. ORMS B elevator showed up John Toczek is the assistant vice president A first on the ground of Predictive Modeling at Chubb in the A Decision Analytics and Predictive Modeling C floor. A B C ??
Table 1.
Question: For scenario No. 11, which elevator
Journals INFORMS PubsOnLine Suite is your personal link to all areas of OR/MS research and discoveries giving you a serious competitive advantage. Subscribe to INFORMS PubsOnLine Suite and save!
SUBSCRIBE TODAY!
Contact us at 1.800.4.INFORMs, 443.757.3500, or informs@informs.org. Subscribe to all 14 INFORMS Journals Online for only $109! Members-only benefit: Access all journal issues back to vol. 1. no. 1
department. He earned his BSc. in chemical engineering at Drexel University (1996) and his MSc. in operations research from Virginia Commonwealth University (2005).
ACCESS ALL JOURNALS FOR
$109
Pubs
www.informs.org/members-pubssuite
14 | ORMS Today
|
June 2016
ormstoday.informs.org
DATA ANALYTICA CEREBRUM understanding the underlying methodology and mindset of how to approach and handle data is one of the most important things analytics professionals need to know. informS intensive classroom courses will help enhance the skills, tools and methods you need to make your projects a success.
SAMPLING BIAS PRESCRIPTIVE PREDICTIVE STOCHASTIC MODELS NON-TECHNICAL DECISION MAKERS
UNSTRUCTURED PROBLEMS
REGRESSION OPTIMIZATION vs. SIMULATION DISPARATE INFORMATION
UPCOMING CLASS:
essential practice skills for high-impact analytics projects June 16–17, 2016 university of denver, ricketson law buildingsturm college of law, room 370 2255 e. evans ave., denver, co 80208
Register at www.informs.org/continuinged
CHART1
Newsmakers
Salary survey, IAAA & award-winning app Report: Data scientists command big bucks The demand for experienced data scientists continues to far outpace the supply, yet the base salaries for most levels of data scientists remained steady or had only modest gains (1 percent to 3 percent) compared to the previous year, and those at the very top of the pay scale (level 3 managers) saw a 4 percent drop according to a recent report.The “Burtch Works Study: Salaries of Data Scientists 2016” involved 374 confirmed data sciences, 69 percent of whom were identified as “individual contributors.” “Managers” comprised the remaining 31 percent. Not to worry; data scientists of all ranks remain well compensated. According to the report, the median base annual salaries of data scientists ranges from $97,000 to $240,000, depending on a long list of factors including experience, education, industry, location and job responsibility. And that doesn’t include bonuses. The report notes that compared to other predictive analytics professionals, data scientists earn higher median base salaries across every job category. The difference in base salaries is largest among individual contributors, where data scientists earn from 22 percent to 39 percent more than other predictive analytics professionals. For example, individual contributors at level 1 earn a median base salary of $97,000 within data science, compared to $76,000 in other predictive analytics positions. To download a free copy of the complete report, visit: http://www. burtchworks.com/big-data-analystsalary/big-data-career-tips/ Innovative Applications in Analytics Award The Innovative Applications in Analytics Award (IAAA), which recognizes creative and unique developments, applications or combinations of analytical techniques used in practice, has garnered significant interest in its brief life.The IAAA for 2016 was 16 | ORMS Today
|
June 2016
awarded to an MIT-led team for its submission entitled, “An Analytics Approach to the Clock Drawing Test for Cognitive Impairment,” during the recent INFORMS Conference on Analytics & Operations Research in Orlando. Fla. The Innovative Applications in Analytics Award is the flagship competition of the Analytics Society of INFORMS.The purpose of the award is to recognize the creative and unique application of a combination of analytical techniques in a new area.The prize promotes the awareness and value of the creative combination of analytics techniques in unusual applications to provide insights and business value.
gle Maps, a quality index average, and – for the state of Georgia – the major insurance providers accepted. Prashant Tailor, a recent graduate of the Stewart School of Industrial & Systems Engineering (ISyE) and a member of the FindED team, served for several years as an undergraduate researcher for ISyE professor and INFORMS member Eva Lee. It was through his work with Lee in support of her research that Tailor came up with the initial idea for FindED. “I was working at Grady and other hospitals,” Tailor explains, “and I noticed that there was an excessively long wait, especially at Grady. I thought,‘I can look up wait times for other things online – like food – why can I not do this for a hospital?’” Tailor emphasizes the importance of the team’s interdisciplinary nature as integral to the success of putting FindED together. In addition to Tailor, the co-founders include Farhan Khan (computer science major), Dale Rivera (computer science) and Tony Shu (materials science & engineering and computer science). “The IE aspect to me is how can I look at the entire system? How do I use a data set to model a real-life system? That’s pure ISyE,”Tailor said. T h e t e a m ’s p ro j e c t g a r n e re d considerable attention, including a front-page appearance on Reddit and a first-place finish in the technical paper competition at the 2016 IIE Southeast Regional Conference. “The key challenges are to ensure that the app is relevant, the information provided is objective and up-to-date, and that patients can choose what matters to them the most – how long they have to wait versus quality ranking, or insurance acceptance, etc.” says Lee, the team’s advisor. “We have already beta-tested its usability to over 60,000 users. The win definitely motivates broader dissemination.” ORMS
Students’ app finds shortest wait time at hospital EDs Numerous studies have shown that patient demand on hospital emergency departments (ED) has increased exponentially, and that demand grows each year. From 2003 to 2009, the average wait time in U.S. EDs between arrival and being seen by a medical professional increased 25 percent, from 46.5 minutes to 58.1 minutes. Prolonged wait times are reported to be a central concern in EDs and are a major reason why patients leave the ED without being seen. Until recently, there has not been a tool that helps would-be ED users – specifically those who do not arrive via ambulance – determine which nearby hospital has the shortest wait time. Thanks to an interdisciplinary undergraduate team of students from Georgia Tech, the web-based application FindED does just that. FindED users are taken to a screen that displays hospitals within a 15-mile radius of the user’s location. Each hospital shows the wait time (based on an annually reported average), the travel time based on real-time FindED team members (l-r) Farhan Khan, Dale Rivera, Tony Shu information from Gooand Prashant Tailor. ormstoday.informs.org
What’s Your StORy? John Toczek AVP Predictive Modeling & Analytics at Chubb INFORMS member since 2005 What has been your best INFORMS experience so far? I love the Business Analytics conferences and have been to all of them in the past 10 years. The talks are always great but the camaraderie is what I look forward to most. Some people you only see once a year but I’ve made some great friendships through those interactions. There is a ton of excitement for our field and I’ve been fortunate to have a front row seat to witness the growth. How did you become “the PuzzlOR”? I blindly sent an email to Peter Horner (Editor, OR/MS Today & Analytics) in January 2008 that said, “I've been kicking around an idea for a new section in OR/MS Today called the PuzzlOR. It would be a relatively easy problem that requires OR techniques to solve.” Peter replied, “We've never published anything like this, but I think it's crazy enough to work.” It struck a chord with the greater OR community and I think that is because when you distill what we do down to its essence, we all love puzzles. It’s hard to believe I’ve been writing these for 8 years now. What recent project have you been involved in that you are proud of? There are so many exciting projects going on right now in the Global Analytics Predictive Modeling group at Chubb insurance. I’m in the middle of a project to predict automobile accident fraud in Mexico. And I’m working on some rate modeling for a life insurance project in Thailand. Did you know you can buy life insurance at car dealerships there? The cultural dimensions add another level of fascination when building and implementing models. Here’s a puzzle for you: There are a dozen eggs in a carton. Twelve people each take a single egg, but there is one egg left in the carton. How? There was a tiny chicken in the carton and she laid a 13th egg while the other 12 were being taken. (Either that or it was a baker’s dozen.) If you have a different answer for this puzzle, comment on the INFORMS connect post!
More questions for John? Ask him in the Open Forum on INFORMS Connect!
http://connect.informs.org
The M1 Abrams tank entered service in 1980. Does the U.S Army need a newer version? Source: Department of Defense
Reforming America’s
defense procurement Hawks and doves agree U.S. military spending needs a serious, systematic review, but widespread conflicts of interest present a major challenge. By Douglas A. Samuelson 18 | ORMS Today
|
June 2016
Hawks and doves can agree on one conclusion: U.S. military spending needs a serious, systematic review.The world’s dominant military power is also, by far, the biggest spender on defense-related acquisition and activities.The United States is also a major supplier of arms and training to other countries. But is the United States buying the right capabilities, in the right amounts, and in the best way to avoid waste and abuse? Many knowledgeable people think not. This issue isn’t about well-publicized claims that the military paid several hundred dollars each for hammers and toilet seats and coffeemakers. Some of these stories came from misinterpretations of accounting; a major contractor was instructed, part way through a big system procurement, to show its overhead costs by budget line item. Having made no such accounting setup from the start, they simply took their calculated overhead cost and divided it by the number of ormstoday.informs.org
line items.Thus both the hammer and the airframe got allocated about $400. The coffeemaker, on the other hand, really did cost several hundred dollars. Pilots flying long missions find it very helpful to be able to make coffee in flight. However, as one veteran Air Force pilot dryly told this reporter, “Your basic Mr. Coffee doesn’t hold together very well in a four-g turn, and getting coffee sprayed all over yourself while flying isn’t desirable.” Of course some items cost more than they should. But the bigger issue is whether certain items should be bought at all, and if so, how many. Do we need a newer version of the Abrams Is the F-35 joint strike fighter a good all-purpose fighter and attack aircraft, or an expensive boondoggle? tank? Do we need to buy more Source: Department of Defense C-17 cargo aircraft than the Pentagon says it needs? Do we need a new strategic bomber? Is the F-35 joint strike all too accurate, with terrible effects on U. S. foreign Both fighter really a good all-purpose fighter and attack policy and domestic spending priorities. aircraft, or an expensive boondoggle that serves none The book documented a number of major of its intended purposes well? systems procurements in which suppliers had Focusing on cost-cutting actually appears to obstrong interests in pushing for certain systems scure more important questions. Frank Kendall, unand for making more of them after the Pentagon and dersecretary of defense for acquisition, technology believed the United States had enough. These and logistics, recently presented an analysis of cost companies make large campaign contributions have such a overruns and management errors over the past 30 and employ large numbers of lobbyists. They years. He concluded that procurements tend to be distribute production sub-tasks among numerous less well managed and produce worse results during Congressional districts, so many representatives times of tight budgets. “When funds are tight,” he and senators stand to gain jobs for their districts by explained, “people tend to be more optimistic about having production continue. to initiate and what can reasonably be done. Procurements run In some cases, the McCartneys went on to without severe budget constraints tended to be run argue, both military leaders and industry have more realistically.” He also noted that good people such a strong incentive to initiate and continue are vital to having programs run well, and recruiting large prog rams that they slant intelligence and retaining the best people is more difficult in lean estimates to overstate threats the country faces. times. Then the possession of new weapons creates an incentive to use them and to encourage other that they A More Critical Viewpoint countries to do so. Some critics go further. As a reporter for the Chicago This is obviously a strong and controversial Daily News and then Knight Ridder Newspapers for point of view. However, this book, paradoxically, nearly 40 years, Jim McCartney was deeply versed will be of particular interest and value to readin military and related industrial affairs. He covered ers who disagree with its main conclusion – that estimates to military spending for nearly 30 years and eventually is, readers who support continuing and growing wrote most of a book about it: “America’s War MaU.S. projection of force and influence. The auchine:Vested Interests, Endless Conflicts.” His widow thors presented compelling and detailed evidence completed the book after he passed away in 2011. In of collusion and lobbying by major defense conit, they made a cogent case that President Eisenhow- tractors to influence not only procurement, but the country faces. er’s warning, in his farewell address, about the influultimately also the military capabilities the U.S. ence of a growing “military-industrial complex” was will have and hence its readiness to commit to
military leaders industry
strong incentive
continue large programs
slant intelligence overstate threats
June 2016
|
ORMS Today
| 19
Defense Procurement
Production of the A-10 Thunderbolt has been discontinued. Is a new version on the way? Source: Department of Defense
proposed courses of action. When the mix of capabilities, a vague policy objective and military careerism intersect, military leaders tend to push policy toward conflicts that will advance careers by providing combat experience without risk of large-scale defeats. In such a climate, industry is easily induced to produce more of what the current leaders know and like. Indeed, large industrial firms mount intensive lobbying efforts not only to help them win contracts, but also to influence what kinds of procurements are chosen – and hence what conflicts the United States will regard as acceptable. Influences on Policy The idea that lobbying over procurement and career advancement can end up wrongfully influencing policy is neither a new nor a radical point of view. H. R. McMaster meticulously documented just such an assessment of the decisions that led to the United States war in Vietnam, partly in response to what he saw as a whitewash in former Secretary McNamara’s retrospective. His critique of a culture of self-serving decisions and slanted intelligence assessments by senior military leaders and their civilian superiors, 20 | ORMS Today
|
June 2016
pushing the country toward the wrong war approached in the wrong way, was scathing. That he remained in the Army and is now a lieutenant general is testimony to the persuasiveness of his reasoning as viewed by his fellow officers. In addition to McMaster’s analysis of careerism and self-interest, some observers might suggest that the Vietnam debacle may have been made more likely by the shift in approach, early in the Kennedy administration, away from an emphasis on massive nuclear deterrence toward more capability to fight small wars. The relationship between what leaders think we should be able to do and what we end up doing seems to call for more thinking – and thinking not overly influenced by the few big firms that dominate procurement. Following the McCartneys’ line of thought, it is clear that the current decision-making system will tend to favor missiles and airplanes over such promising approaches as village stability operations (VSO), described earlier this year in Analytics magazine [6], because VSO doesn’t generate production jobs in numerous congressional districts – and because the results take longer and are harder to see. ormstoday.informs.org
Similarly, this reporter has heard harsh complaints from Army and Marine commanders about the Air Force’s long-expressed desire, now fulfilled, to discontinue producing the A-10 assault and observation aircraft. This preference by the Air Force came about at least in part because its pilots would rather fly airplanes better suited to air-to-air combat and less suited to the A-10’s strength – observation to aid and support artillery and ground forces. This reporter has also heard assertions by Air Force analysts that this shift indicates a growing belief that future engagements should rely more on air power and less on ground forces – a shift in tactics closely related to a shift in available resources if ever there was one. Note that this shift in approach does not appear to have involved much assessment of which tactics are actually most effective. New Missions and Threats Now, however, there is the additional issue that new missions and threats may require technology quite different from the big firms’ expertise, so the current procurement system tends to be biased against needed innovation. Even defining what the missions and threats of greatest interest might be takes the conversation in directions away from what it has been – and hence away from the currently identified and trusted expertise. If new approaches are needed, are the masters of older approaches the best parties to evaluate which new approaches are good? If not, then who? As Michelle Flournoy, a co-founder and current CEO of the Center for a New American Security who also served as undersecretary of defense for policy, 2009-2012, conceded in response to a question at an event to launch a new report on U.S. strategy, “How to involve newer, smaller providers and promote innovation is a big ongoing concern, and nobody really knows how to address it effectively.” Again, cost-cutting concerns add to the difficulty. In recent years, many contracting offices have moved away from support contracts focused on providing good analysts to support multiple tasks over several years, to contracts that have vendors compete task by task. This pushes vendors to bid the minimum qualifications needed for each task, pushing out the most experienced people and disrupting continuity. On the other hand,“Just keep going back to the people who did well before” can be either a wise approach to making use of acquired knowledge or another formula for stifling innovation. Different contracting approaches are needed, and some senior officers and managers are grappling with the issue – but this looks like a good opportunity for systematic analysis, if someone can develop and support a way to identify and compensate the right people to do it.
Whether the United States fights too many wars, and the wrong ones, is one issue; whether it is preparing properly for the necessary missions (including humanitarian relief, a major component of the U.S. military’s activities) is another. The McCartneys’ book did not grapple with the latter issue but provides excellent background to do so. Humanitarian relief and stability operations utilize different resources from either conventional or asymmetric combat. For example, one reason for the bad outcome after the quick victory in Iraq in 2003 is that the optimization of logistical activity in support of combat operations had left the United States and its allies short of trucks, among other resources, to support restoring civilian services. It has also been well argued previously (Elliot Cohen’s book is one of the most widely praised examples) that civilian leaders need to pay less attention than they usually do to how the military might carry out a mission and more to how the military leadership frames its analysis of what to recommend. The McCartneys’ book provides much information to support challenges to the usual assumptions – again with the caveat that one need not agree with their main point of view to find the information valuable. At least the discussions would be worth having. In short, the development of many entrenched interests is complicating both the assessment of national needs and the procurement of the means to address those needs. OR/MS analysts who have taken the time to become familiar with the details of these issues should find no shortage of opportunities to apply their expertise for the good of the country. Finding a receptive audience at the senior policy levels, however, may prove to be the biggest challenge of all. ORMS
Whether the
United States fights
too many wars, and the
wrong ones, is one issue; whether it is
preparing properly for the
necessary missions is another.
Douglas A. Samuelson (samuelsondoug@yahoo.com) is president and chief scientist of InfoLogix, Inc., a small R&D and consulting company in Annandale, Va.
REFERENCES 1. Center for a New American Security, “Extending American Power: Strategies to Expand U.S. Engagement in a Competitive New World Order,” May 16, 2016. Available online from www.cnas.org 2. Elliot Cohen, “Supreme Command: Soldiers, Statesmen and Leadership in Wartime,” Free Press, 2002. 3. James McCartney with Molly Sinclair McCartney, “America’s War Machine: Vested Interests, Endless Conflicts,” Thomas Dunne Books, St. Martin’s Press, 2015. 4. H. R. McMaster, “Dereliction of Duty: Lyndon Johnson, Robert McNamara, the Joint Chiefs of Staff, and the Lies that Led to Vietnam,” Harper, 1996. 5. Robert S. McNamara and Brian VanDeMark, “In Retrospect: The Tragedy and Lessons of Vietnam,” Vintage, 1996. 6. Douglas A. Samuelson, “Changing the Game: How to Defeat Violent Extremism,” Analytics, January-February 2016. 7. “The State of Defense Acquisition,” Center for Strategic and International Studies (CSIS) briefing by Frank Kendall, May 10, 2016; may be viewed online at www.csis.org/events/statedefense-acquisition
June 2016
|
ORMS Today
| 21
UPS optimally routes 55,000 drivers, each of whom makes on average more than a hundred deliveries a day. Source: UPS
Edelman Award
‘ORION’ delivers success for UPS Revolutionary routing system boosts driver efficiency, cost savings, customer service and the environment.
By Peter Horner
22 | ORMS Today
|
June 2016
UPS was faced with what amounted to the Mother of all Traveling Salesman Problems in its Small Package Operations: how to optimally deploy 55,000 drivers based across the United States, each of whom makes on average more than a hundred deliveries a day. To make the problem even more interesting, UPS added a largely safety-driven constraint: minimize the number of left-hand turns on each driver’s route. The goal: boost cost savings, driver efficiency, production and safety, environmental friendliness and customer service all at the same time by optimizing driver routing and reducing the total number of miles driven and fuel consumed throughout its U.S. pickup and delivery system. That’s a tall order, but UPS delivered thanks to ORION (On-Road Integrated Optimization and Navigation), a 10-year-long project developed by the UPS Operations Research group that revolutionized the company’s pickup and delivery (P&D) operations, ultimately met and exceeded its goals, and earned UPS the 2016 Franz Edelman Award from INFORMS. The prestigious Edelman Award, considered the “Super Bowl” of O.R. practice, was presented at an Oscar-like gala held in conjunction with the ormstoday.informs.org
INFORMS Conference on Business Analytics & Operations Research in April in Orlando, Fla. “ORION is a testament to the scale of innovation that can be accomplished when operations research, information technology and business processes are seamlessly integrated,” says Juan Perez, UPS chief information officer. “Success didn’t come overnight. It took hard work by 700-plus UPS employees who dedicated themselves to the development and deployment of ORION for many years. And we are just at the beginning of harnessing the power of ORION to fuel new, innovative products and services.” The Project As the prize-winning UPS team outlined in its presentation, ORION sits atop a package flow technology (PFT) foundation that UPS developed to streamline and modernize its P&D operations. Launched in 2003, PFT combines data from multiple sources (public as well as proprietary) and advanced analytical tools to provide UPS with “unparalleled flexibility and efficiency.” A decade later, Information Week named PFT one of the “20 Great Ideas to Steal.” PFT was explicitly built to support the use of advanced optimization in planning and execution of its P&D operations. However, the project’s initial route optimization algorithms were difficult to implement in practice. As noted in the presentation, UPS “went back to the drawing board and had to rethink and relearn everything it had known about creating effective and efficient routes. It had to blend its 108-year-old practices with 21st century technology.” The result was a field-tested ORION that uses advanced fleet telematics and complex algorithms to crunch thousands of pages of code and more than 250 million address data points to provide UPS drivers with optimized delivery routes. As of December 2015, ORION was being used by 35,000 UPS drivers in the United States, and every morning they receive an optimized sequence in which the (pre) assigned packages are delivered. Full deployment to 55,000 drivers is expected by the end of this year. According to UPS, ORION saved the company $320 million by the end of 2015, but the biggest savings are going forward. Once fully deployed, ORION is expected to account for an estimated $300 million to $400 million in annual savings and cost avoidance thanks to 100 million miles less traveled, 10 million gallons of fuel not consumed and a reduction of 100,000 metric tons in CO2 emissions per year, not to mention a significant increase in deliveries per driver per day. Add it all up and it’s a nice annual return on a $250 million total investment. “ORION has been a game changer for UPS,” says Mark Wallace, senior vice president of global
Five other Edelman finalists Along with UPS, this year’s Franz Edelman Award competition finalists included: • 360i for “360is Digital Nervous System” • BNY Mellon for “Transition State and End State Optimization Used in the BNY Mellon U.S. Tri-Party Repo Infrastructure Reform Program” • Chilean Professional Soccer Association (ANFP) for “Operations Research Transforms Scheduling of Chilean Soccer Leagues and South American World Cup Qualifiers” • New York City Police Department (NYPD) for “Domain Awareness System (DAS)” • U.S. Army Communications-Electronics Command (CECOM) for “Bayesian Networks for U.S. Army Electronics Equipment Diagnostic Applications: CECOM Equipment Diagnostic Analysis Tool, Virtual Logistics Assistance Representative”
engineering and sustainability at UPS. “ORION has made us better at serving our customers – how they want, when they want, where they want.”
“Thousands of people
The Prize-Winning Team The strategic importance of ORION to UPS and the fact that the company’s appreciation of operations research and its O.R group goes all the way to the very top of the organization was underscored by CEO David Abney flying into Orlando for the Edelman competition and accepting the award on behalf of UPS at the dinner gala. “This is another example of a group taking an excellent idea and then transforming that idea into a game-changing technology,” Abney told the packed ballroom.“Operations research has been a key to our success for the 42 years that I’ve been at UPS, and I’m sure it will always continue to be that way.” Abney indicated that UPS will donate its prize money from the Edelman competition to help support students who plan to pursue a career in operations research. Headquartered in Sandy Springs, Ga., UPS is a global leader in logistics, offering a broad range of solutions including the transportation of packages and freight, the facilitation of international trade and the deployment of advanced technology to more efficiently manage the world of business. UPS serves more than 220 countries and territories worldwide. No doubt every seasoned O.R. analyst has worked on a technically sound and innovative O.R. and analytics-driven project that failed to live up to expectations due to a lack of buy-in from key constituents.The ORION project obviously had the support of UPS’s top management, but what about its 55,000 U.S.-based drivers, many of whom had driven a particular route a particular way for years? How would they respond to daily, computer-generated “optimized” routes that veered away from their routine? The Edelman-winning presentation featured a dramatic video of UPS employees, from C-level
now have
June 2016
O.R. tools in their hands to
make better decisions.” – Jack Levis
|
ORMS Today
| 23
Edelman Award executives to drivers, extolling the virtues of operations research and the ORION project. As one driver puts it, “[ORION] tells you exactly how to drive to your first stop and how to get back to the building from your last stop, so I can stay focused on safety aspects and my customers, addressing them properly, seeing how they’re doing and just being professional,” he said, adding that he used to do 110 stops a day and now does 115 to 118 stops in the same amount of time. Chuck Holland, vice president of industrial engineering, and Jack Levis, senior director of process management and a longtime, active member of INFORMS, were prominent members of UPS’s Edelman presentation team. Both emphasized the teamoriented approach to the project that CEO Abney referenced, including the critical role played by drivers. “By the end of the year we’ll have 55,000 drivers using ORION, 700 people in the United States deploying it, thousands of operations managers taking advantage of it, let alone the great operations research team that developed the algorithms and the technology behind it,” Holland said. “At UPS we really consider ourselves a team, and I think this is just a perfect example of teamwork throughout the company. “The first few attempts at the algorithm did not work well in the field,” Holland continued.“We took a step back, and we worked very closely with our drivers. In the presentation today, Jack [Levis] made a comment
UPS’ Operations Research group, led by Jack Levis (front, center) developed ORION over a 10-year period. Source: UPS
about ORION thinking less like a computer and more like a driver. We were able to do that; we were able to gain great acceptance from our drivers. If you ask them, the vast majority will say it takes a good amount of pressure off of them so they can devote more time
General Motors awarded INFORMS Prize General Motors Corp., which is using big data and advanced analytics to predict failure of certain automotive components and systems before customers are affected, was named the winner of the 2016 INFORMS Prize for operations research and the management sciences. This year’s INFORMS Prize was presented at the 2016 INFORMS Conference on Analytics & Operations Research in Orlando, Fla. The INFORMS Prize honors effective integration of operations research in organizational decision-making. The award is given to an organization such as GM that has repeatedly applied the principles of O.R. in pioneering, varied, novel and lasting ways. Unlike the Edelman Award, which honors an impactful project, the INFORMS Prize salutes successful and sustained integration of O.R. and analytics throughout an organization. Industry-first proactive alert messages sent to customers through GM’s OnStar system covering potential issues with a vehicle’s battery, fuel pump or starter can transform an emergency repair into planned maintenance. A recent example of applying operations research and management science to the most complex issues the company faces led to the INFORMS Prize. “Over the last seven decades, OR/MS techniques have been used to improve our understanding of everything from prognostics to traffic science and supply chain logistics to manufacturing productivity, product development and vehicles telematics and prognostics,” says Gary Smyth,
24 | ORMS Today
|
June 2016
executive director of GM Global R&D Laboratories. “These approaches to problem-solving permeate almost everything we do.” G M h a s h u n d r e d s of O R / M S practitioners worldwide who play a vital role in everything from designing, building, selling and servicing vehicles to purchasing, logistics and quality. The team is constantly developing new business models and vetting emerging opportunities. Another example of management science positively impacting the business is helping to understand what products and features customers most want to create, as well as price features and option packages that would sell best. That work extends to determine the ideal number of vehicles and what vehicle variants Chevrolet, Buick, GMC and Cadillac dealers in the United States should stock. In addition, advanced analytics help dealers achieve GM’s goal of creating customers for life. The company recently received the 2015 overall manufacturer loyalty award from IHS. The impact OR/MS is now having to the business can be traced to 2007, when GM created a center of expertise for operations research to promote best practices and transfer new technologies. GM has since expanded to include partner teams in product development, supply chain, finance, information technology and other teams. ormstoday.informs.org
to things that are more important, such as safety and servicing our customers.” Added Levis: “UPS wants to make better decisions and we have.Thousands and thousands of people now have O.R. tools in their hands to make better decisions.Think about it – 55,000 drivers delivering 16 million packages to 8 million customers every day using O.R. What better testament can there be for operations research?” Longtime INFORMS members Randall Robinson and Ananth Iyer served as volunteer coaches for the victorious UPS team.
ORION uses advanced fleet telematics and complex algorithms to crunch thousands of pages of code and more than 250 million address data points.
The Judges The Edelman Award is a nearly yearlong competition that begins with a call for nominations, followed by a vetting and verification process. Once the nominations are culled to six finalists, the competition culminates each spring with presentations before a panel of judges at the INFORMS Conference on Business Analytics and Operations Research. After listening to all of the presentations and questioning the presenters, the judges debated behind closed doors before selecting a winner. So what put UPS over the top in a highly competitive and diverse field of Edelman finalists (see sidebar)? “The integration of the data and optimization, the fact that you needed the two of them to work well together,” said Mike Trick, professor at Carnegie Mellon and chairperson of this year’s Edelman Award Committee, who led the panel of judges. “They had a really complicated implementation and they did it extraordinarily well. This was a 10-year project building off a previous project, so it was a long time coming. With the data being in the state that it was, it’s amazing that they were able to step back and say, ‘No, this isn’t just about optimization. If we don’t get the data right, if we don’t get the underlying rules right, this will never happen.’ The fact that they were able to do that was pretty impressive.” Trick was also impressed by UPS’s ability to garner driver acceptance on a massive scale. “Absolutely, driver buy-in was crucial,” Trick said. “I can just image a driver saying, ‘I’ve been driving this route every day for 20 years, and suddenly you’re telling me a better way of doing it?’The fact that they were able to get that driver buy-in part right is terms of the route ordering is also quite impressive.” Added David Hunt of Oliver Wyman: “It was a very tough decision. There was such a diversity of projects that it was hard to compare them. The one
Source: UPS
With these
thing that I personally really liked about the UPS project is that to me it represented a big scale operations research project in all of its ugliness. I mean they had data issues. They had changed management issues.They had to have extensive training and buy-in by the drivers. It just went through the entire process, and I appreciated that and the effort that went into it.” While each judge has his or her own preferences in terms of what aspect of a project is most important, they all are looking at some combination of technical innovation, obstacles overcome and benefits realized to determine the winner. “The overall challenge that UPS faced to get this project implemented, from taking what was once a simple O.R. algorithm and making it work in practice and deploying it out to an organization with 55,000 drivers, was impressive,” said Irv Lustig, optimization principal at Princeton Consultants. “Getting the math right was a huge challenge, along with working with drivers and people at UPS facilities, handling the changing management issues, and changing the metrics of how they measured success. All of these things put together show that with these massive projects, it’s not just about the algorithms; it’s about all the things you have to do around it to be successful.” How important is winning the Edelman Award? Minutes after INFORMS President Ed Kaplan announced the winner, and the audience erupted in applause and the music started thumping, UPS V.P. Holland was asked that very question. His answer: “I postponed my retirement by one year because of this. We weren’t sure if we were going to compete in 2015 or 2016. Once we decided we would wait until 2016, I postponed my retirement in hopes of being part of a celebration just like this.” ORMS
massive projects, it’s not just
about the algorithms; it’s about
all the things you have to do
around it to be
successful.
Peter Horner (peter.horner@mail.informs.org) is the editor of OR/MS Today and Analytics magazines.
June 2016
|
ORMS Today
| 25
Optimizing crop management “Smart� application of fertilizer illustrates payoff in using analytical tools to enhance crop yields and improve the environment. By Joseph Byrum
The modern farm is becoming a proving ground for the value of data analytics. For example, by looking at systems that optimize the management of fertilizer, or nitrogen, we can see that the reward for improved decision-making goes far beyond simple economics. Better run farms produce more food, contributing to global food security. Properly managed nitrogen also happens to be essential to improving the environment and water quality. In short, data analytics are key to a healthier and happier future for millions around the globe. The environmental and food security challenges the world faces in the decades ahead are truly complex, and the job of data analytics is to make sense out of inherently complex systems, like those we find in agriculture. When it comes to growing crops, no one field is the same as another. Natural factors such as soil quality and topography introduce unique characteristics that affect the rate of growth and health of plants grown
For plants like corn, nitrogen tends to be the most important factor for yield.
26 | ORMS Today
|
June 2016
ormstoday.informs.org
t h e re i n . P l a n t s a l s o d e p e n d o n sunlight and water for growth, so the weather introduces a random source of variability. And then there’s the managed source of variation found in the application of fertilizer. Whether the farmer has a small plot in Africa or a large growing operation in Iowa, there are always ways to do things smarter and more efficiently. The tools of operations research are capable of tackling each of the three main sources of variability to improve decision-making. The payoff in using analytical tools to optimize crop management is perhaps best revealed in exploring the application of nitrogen to enhance crop yields. For plants like cor n, nitrogen tends to be the most important factor for yield, but it is not so simple. ApField monitoring documents changes that are naturally occurring within a field. plying more nitrogen will not always Source: Syngenta produce better results. To the contrary, an overabundance of nitrogen is a data to select the best crop genetics, inputs and Whether the known menace to the environment. This is why growing techniques that are a match for the field one-size-fits-all approaches to growing are beconditions and weather. coming obsolete. We see the highest yields coma small plot Achieving Optimization ing from customized approaches that use hard Achieving that optimization requires better measurement, and this poses a practical dilemma. Farmers must invest in an array of remote sensors and or a related analytical tools to achieve the benefits of better management, yet some farmers prefer to stick to tried-and-true methods.They base how much nitrogen they will use this year on how much was used last year. They are more likely to invest more capital on a new tractor or combine – the value of this equipment is well known – than to spend on sensors in Iowa, and software that are unfamiliar to them. The tools themselves hold the answer to whether such an investment makes economic sense. Sensors are an integral component of yield monitors that measure field performance against ways to do revenue. Simply put, more precise measurements enable farmers to make smarter, more accurate decisions that have an impact on their bottom line. Properly implemented sensor technologies are critical to achieving the data density required and to ensure farmers have the economically actionable intelligence they need to improve their management practices. The marketing of these critical tools must be optimized to reflect the practical needs of farmers. Remote sensors provide an objective assessSource: Syngenta ment of a plant’s health by measuring chlorophyll
farmer has in Africa
large growing operation there are always things smarter
more efficiently.
June 2016
|
ORMS Today
| 27
Agriculture Analytics
Failure
levels – the greener the plant, the healthier it is. Infrared detectors peer deeply into the plant to detect crop stress, such as the presence of pests, the lack of water and the lack of nutrients. When combined with the technique of infield reference strips, these sensors arm growers with the data they need to more precisely apply nitrogen. If the output of a field matches the yield of the reference strip, no more nitrogen is needed. Conversely, if output is down in comparison, more nitrogen may be needed. Failure to apply nitrogen with precision causes significant harm. As much as nitrogen delivers a massive boost to corn yield, it has an even greater effect in promoting algae growth when fertilizer runoff hits a stream or lake. Eutrophication is the term used to describe the resulting overabundance of nutrients in a body of water. While plankton and algae feast upon the bounty of nitrates, they also multiply rapidly and disrupt the ecosystem’s balance. The algae that die end up consuming enough of the available oxygen that native fish suffocate.
to apply
nitrogen with precision causes
significant harm.
In addition to this, nitrates making their way into the water supply raise significant human health concerns. The Environmental Protection Agency considers levels above 10 parts per million a hazard to drinking water, reflecting an elevated risk of various forms of cancer [1]. The situation is so serious in central Iowa that farmers have their livelihoods at risk in a lawsuit filed by the Des Moines Water Works over runoff. Complex Mathematical Challenge The best way to get ahead of any such developments is to get nitrogen right in the first place, which is to say, by applying no more nitrogen than the plant can absorb. This is a complex mathematical challenge. Nitrogen is soluble in water, so it is swept away by moving water, whether by irrigation or a rainstorm. That means nitrogen levels change rapidly. The job of data analytics is to quantify a plant’s response to varying nitrogen levels. To know how much nutrient to apply to a plant requires an understanding of the plant’s requirements and its ability to absorb them from
Stanford team wins Syngenta Crop Challenge Syngenta and the Analytics Society of INFORMS named Xiaocheng Li, Huaiyang Zhong a n d a s s o c i a te p r o f e s s o r s D av i d L o b e l l a n d S te f a n o Ermon – a team from Stanford University – as the winners of the inaugural Syngenta Crop Challenge in Analytics. T h e te a m wa s awa r d e d a $5,000 prize for its entry, “Hierarchy modeling of s oyb e a n va r iet y yie ld a nd decision making for future planting plan,” which modeled a system for predicting soybean seed variety selection. “It has been a wonderful A team from Stanford University won the inaugural Syngenta Crop Challenge in experience working with Analytics. Syngenta on this project, and we are excited about the impact our work can our findings and proves to us there is a lot of have on improving crop yields and addressing potential for modern operations research and food security challenges,” says Xiaocheng Li. computer science techniques in agriculture.” “Operations research and advanced analytics T h e C h a l l e n g e t a s ke d p a r t i c i p a n t s to can contribute to variety development and d eve l o p a m o d e l t h a t p r e d i c t s t h e s e e d evaluation, re ducing costs and improve d varieties farmers should plant next season to ef ficiency. Ex tracting useful insights from ma ximize yield. The inaugural competition massive, unstructured data sets informed aimed to address the challenge of global food
28 | ORMS Today
|
June 2016
ormstoday.informs.org
the soil, with the economic cost factors always kept in mind. Field experiments are used to test the effectiveness of different levels of nutrient application, but these tests must keep in mind that variability in the fertility of the soil must be factored in for the results to be accurate. The analytics opportunity is to monitor in space and time in an effort to document the changes that are naturally occurring within a field. Statistical methods that rely on history alone are inadequate; experiments are necessary. For nitrogen optimization, treated sites and untreated sites would be established. Spatial statistics can be used to analyze covariance in the experimental sites, ultimately allowing the farmer to more precisely interpret the trends. Instead of guessing what is going on, he will know what is happening. By acting on solid information rather than intuition, his probability of success will increase because it will no longer be a matter of luck. Farmers who have yet to explore the use of data analytics and sensor technologies are going
to have to dive into these technologies to stay competitive in the years ahead. The Iowa Soybean Association keeps track of the performance of nitrogen sensing in the field. Most farmers are reporting savings of between $10 and $20 per acre in reduced fertilizer costs [2]. In many cases, growers recoup the cost of sensors within a year or two. At the same time, the industry also needs to step up to the challenge to implementing easy-touse data analytics tools. That is a necessary step in fulfilling the promise of nitrogen optimization and contributing to global food security and a cleaner environment.ORMS
security by fueling innovation among experts applying advanced analytics in biochemistry and agriculture. “Global food security is one of the greatest challenges facing the next generation, and there is a significant need to engage a broader talent base into agriculture,” says Joseph Byrum, Syngenta head of soybean seeds product development and lead for the Syngenta Crop Challenge in Analytics Committee. “This competition clearly demonstrated that people outside and adjacent to the industry can make noteworthy contributions.” The finalists made their presentations at the INFORMS Conference on Business Analytics & Operations Research in Orlando, Fla. Programs were evaluated based on the rigor and validity of the process used to determine seed varieties, the quality of the proposed solution and the finalists’ ability to clearly articulate the solution and its methodology. The runner up, “Decision assist tool for seed variety selection to provide best yield in known soil and uncer tain future weather conditions,” authored by Nataraju Vusirikala, M e h u l B a n s a l a n d Pr a t h a p S i va K i s h o r e Kommi, received a $2,500 prize. The third place entry, “Balancing weather risk and crop yield for soybean variety selection,” authored by
Bhupesh Shetty, Ling Tong and Samuel Bure, received a $1,000 prize. “The submissions from the Syngenta Crop Challenge in Analytics represent best in class science,” Byrum adds. “What is striking is the overall professionalism, quality and effort that the finalists put into the presentations. The teams were clearly committed and had a deep connection to the challenge.” Syngenta, a global agribusiness headquartered in Switzerland, donated the prize money from its 2015 Franz Edelman Award win in support of a commitment to run the Syngenta Crop Challenge for the next four years. “Syngenta is a great example of a company using operations research to better both its own performance as well as to help better society,” says Melissa Moore, executive director of INFORMS. “In 2015 Syngenta won the Franz Edelman award for using operations research and analytics to make better breeding decisions to reduce the time and cost required to develop crops with high productivity. Their effor ts, including the Crop Challenge in Analytics, are putting them at the forefront of utilizing operations research to transform the agriculture industry.”
Joseph Byrum, Ph.D., MBA, PMP, is senior R&D and strategic marketing executive in Life Sciences – Global Product Development, Innovation and Delivery at Syngenta.
REFERENCES 1. http://www.ncbi.nlm.nih.gov/pubmed/11338313 2. http://www.cals.uidaho.edu/edComm/pdf/BUL/BUL896.pdf
For more details about the Syngenta Crop Challenge and to register for the 2017 Challenge, visit www.ideaconnection. com/syngenta-crop-challenge.
June 2016
|
ORMS Today
| 29
P ∈ OR
(P-values in operations research) …M N O P Q R S T…
You don’t need a license to download R, but you should have a good understanding of p-values.
By Scott Nestler and Harrison Schramm
30 | ORMS Today
|
June 2016
Several of our colleagues asked, “Why is this article necessary? Shouldn’t the American Statistical Association’s (ASA) statement on p-values be enough?” Our answer to this is affirmative; the ASA’s statement is one of the best written documents we have seen this year and nothing we are going to contribute here should be seen as a substitute for reading the source article in its entirety. Our purpose is to draw attention from the O.R. community to an issue that impacts us every day in practice, but that we may have missed in our finite journal attention spans. It is of particular importance to our community, which tends to have strong computer programming skills and a “can do” attitude. In summary, you don’t need a license to download R – but you should have an understanding of what statements such as p-value: .096 really mean.That understanding should be deeper than simple comparison with a threshold value. If you look for “O.R.” in the alphabet you find the letter “p” squarely in the middle.This circumstance of language happens to reflect that statistical hypothesis testing, based on a “p-value,” is central to much of what we do. Many of the advanced analytic methods employed by operations researchers are based in, or draw on, statistical analyses. It is therefore no coincidence that the first content chapter of Morse and Kimball’s “Methods of Operations Research” (1946) is devoted to probability. The p-value concept was introduced in the late 18th century by the great Pierre Laplace – the father of many transformative ideas. Formalization by Karl Pearson and advocacy by Ronald Fisher in the early 20th century (e.g., “The Lady Tasting Tea” experiment) led to establishment of the p-value as a workhorse of inferential statistics and the notion of .05 (or 1 in 20) as a commonly accepted surrogate for statistical significance [1, 2]. The advantage – and disadvantage – of the p-value is that it reduces complex hypothesis tests to a single diagnostic value, and this value is “level” across diverse methods; we interpret the meaning of the p-value the same way regardless of the underlying mathematics. Earlier this year, the ASA published “ASA Statement on Statistical Significance and P-values,” the culmination of a two-year endeavor [3]. The use of p-values is an issue of statistical practice The p-value is a workhorse of inferential as opposed to theory. It affects each of statistics. us in at least one way or another—as ormstoday.informs.org
12 10 8 y
2 0
2
4
6
8
10
6
8
10
yy
10
15
x
5
Our statistics professors were strict adherents to the so-called (by ASA) “Bright-line” method of hypothesis testing, which states: 1, Choose a value for significance a (alpha). 2. Compute p. 3. If p < a, reject the Null Hypothesis.
4
Definitions Matter What’s in a name? Many of us can likely recite the definition of a p-value from a stats course we took at some point along the way. Our response, if put on the spot, would be, “a p-value is the probability of observing a result this extreme or more, given that the null hypotheses is true.”The official but informal definition (per the ASA) is: The probability under a specified statistical model that a statistical summary of the data (for example, the sample mean difference between two compared groups) would be equal to or more extreme than its observed value.
6
a teacher, a student, a researcher or a practitioner, or simply as a citizen who consumes the products and policies that are evaluated by it. In addition to the statement itself, the published version in The American Statistician (TAS) includes commentary from more than two dozen contributors that is also worth reading.
There was explicitly no allowance for the “amount” of difference between p and a consideration of the underlying process from which the data was collected, or ramifications of the decision to be made. The ASA has recommended that we reconsider this procedure. Six Principles for Proper Use and Interpretation of P-Values The most important part of the ASA statement is the listing and subsequent discussion of six principles. In the following section, italicized words are the verbatim principles as they appear in the ASA statement; bold highlights are our attempt to emphasize key points in the principles; and normal text contains our commentary. 1. P-values can indicate how incompatible the data are with a specified statistical model. As the authors indicate, the (in)compatibility identified by a p-value is between the data and the null hypothesis (something like “the population means are all equal”) of a specific model, IF that null hypothesis and any supporting assumptions actually are true. 2. P-values do not measure the probability that the studied hypothesis is true, or the probability that the data were produced by random chance alone. While some students in introductory statistics courses make the
2
4 x
Figure 1: Two data sets exhibiting p=.0499 (top) and p = .0501 (bottom) for simple linear regression. It is obvious why the data set on the (bottom) has a problem (outlier at x = 5); it is not obvious why the data set on the (top) should be “acceptable.”
error mentioned in the first part of this principle, more experienced purveyors of statistical knowledge (authors included) have been known to make the second error. The ASA authors sum this point up as follows: “The p-value … is a statement about data in relation to a specified hypothetical explanation, and is not a statement about the explanation itself.” Some rumination may be advisable here. 3. Scientific conclusions and business or policy decisions should not be based only on whether a p-value passes a specific threshold. Of all the stated principles, this is the one that we think is most deserving of our attention. Making a decision as to whether an effect is “statistically significant” based on some pre-specified level of significance (like a=0.05) is not magical. There is little practical June 2016
|
ORMS Today
| 31
P-value Primer
Ensure
difference between a p-value of .049 and a p-value of .051. Instead of reporting the results of a t-test as t(59)=1.84, p < 0.05, we should actually report the actual p-value, or t(59)=1.84, p =0.0354. Then, allow the person interpreting the result of the test to consider it in terms of practical, rather than statistical significance. 4. Proper inference requires full reporting and transparency. Consider a case study on ethics and data that posits the omission of reporting a data point that does not strengthen the researcher’s argument. Even students in a first year statistics course see problems with that. However, selective use of inference through “p-hacking” is rampant in the literature [4]. The ASA authors emphatically state, “Researchers should disclose the number of hypotheses explored during the study, all data collection decisions, all statistical analyses conducted and all p-values computed.” This will prevent the “Green jelly beans cause acne” finding, as presented in XKCD 882, titled “Significant” [5]. P-hacking is prevalent enough and causes enough concern to have been addressed by a recent 20-minute segment of John Oliver’s “Last Week Tonight” [6]. By “p-hacking” we mean the following (generalized [7]) procedure: UNTIL (p<a DO{ Experiment() } This is more than unethical, it’s just plain wrong; and it erodes the confidence in the public of science.
that we
do not perpetuate improper usage of or
reliance on p-values in the
courses we teach and in
our research and
practice activities.
5. A p-value, or statistical significance, does not measure the size of an effect or the importance of a result. As previously mentioned in principle No. 3, statistical significance and practical (e.g., scientific, human, economic) significance are not the same thing. A smaller p-value does not necessarily imply the presence of a larger effect. As observed with recent “big data” sets, with a large enough sample size, any effect, whether it is really present or not, can generate a tiny p-value [8]. NOTES & REFERENCES 1. http://www.phil.vt.edu/dmayo/PhilStatistics/b%20Fisher%20design%20of%20experiments.pdf 2. http://www.radford.edu/~jaspelme/611/Spring-2007/Cowles-n-Davis_Am-Psyc_orignis-of-05level.pdf 3. http://amstat.tandfonline.com/doi/pdf/10.1080/00031305.2016.1154108 4. http://www.ncbi.nlm.nih.gov/pubmed/22006061 5. http://xkcd.com/882/ 6. http://www.washingtonpost.com/news/speaking-of-science/wp/2016/05/09/john-oliverexplains-why-so-much-science-you-read-about-is-bogus/ 7. You can p-hack at home! Execute the following command in R: shapiro.test(rexp(10)) See how many tries it takes to get the computer to conclude that data from the exponential distribution is actually normal. 8. http://dx.doi.org/10.1287/isre.2013.0480
32 | ORMS Today
|
June 2016
6. By itself, a p-value does not provide a good measure of evidence regarding a model or hypothesis. A p-value is just one of many tools in the quiver of statisticians or other analysts employing statistical methods. While the ASA is not suggesting abandoning the use of p-values, they do propose using them in conjunction with other approaches, such as: confidence, credibility and prediction intervals; Bayesian methods; decision-theoretic modeling; and false discovery rates. Of course, many of these approaches rely on additional assumptions, but they may also more directly get at measuring the size of an effect or the degree of correctness of hypotheses. Implications for Operations Research Professionals The ASA statement should give us pause to consider what it means to operations research analysts, management scientists and analytics professionals. Don’t just take our word for it that this is worthy of your consideration.Take 30 minutes to read the ASA statement in TAS and the associated commentary. Then, consider how to apply the six principles discussed in the statement above. Ensure that we (collectively) do not perpetuate improper usage of or reliance on p-values in the courses we teach and in our research and practice activities. We are not suggesting that p-values be banned, as was done by the editors of Basic and Applied Social Psychology, as they remain a very useful tool, and we both plan to continue using them in our respective practices. Additionally, consider how the ASA statement and principles are interconnected with the issues of reproducibility and replicability. The ASA is also suggesting that statistics may not be as exact a science as many consider it to be. We have now been given permission to answer questions about matters of statistical practice with “it depends” and “too close to call.” These are answers we have previously been culturally hesitant to provide. We wonder how much of this shift in thinking applies to the other sub-disciplines of O.R. This is an open question that we (and we hope you as well) will ponder in the coming months. Please share your thoughts on this subject with our community through INFORMS Connect. We look forward to seeing how this discussion develops. ORMS Scott Nestler, PhD, CAP, PStat, is an associate teaching professor in the newly formed Department of Information Technology, Analytics and Operations (ITAO), Mendoza College of Business, at the University of Notre Dame. Harrison Schramm, CAP, PStat, is a principal operations research analyst at CANA Advisors, LLC.
ormstoday.informs.org
Register Now 2016 INTERNATIONAL CONFERENCE HAWAII June 12–15, 2016 Hilton Waikoloa Village
NETWORK AND SHARE YOUR EXPERIENCES ENHANCED WITH AN AUTHENTIC HAWAIIAN LUAU BY THE BEACH! Come to Hawaii 2016 to listen to an impressive array of over 750 speakers presenting their latest research on: • O.R. in Medicine • Entrepreneurship & Innovation • • • •
Applied Probability Service Operations Big Data Analytics Inventory Systems
• Cloud Computing
• • • •
Practice-Focused Operations Global Supply Chains Mechanism Design & Game Theory Business Strategy
• Robust Optimization • Scheduling and Project Management • and much more
Take this opportunity to network and collaborate with colleagues across the globe from both academia and industry.
PLENARY SPEAKERS: 2016 INTERNATIONAL
HAW II
Qing Wu, Google Data Analytics Frontier in Google/Alphabet Gang Yu, Executive Chairman & Co-Founder of New Peak Group How to Gain Competitive Advantage through Applications of OR/OM
Thomas L. Magnanti,
Singapore University of Technology and Design Contemporary Education for a World that Needs Innovative Products, Services & Systems
REGISTER at http://meetings.informs.org/2016international
Image © David Castillo Dominici | 123rf.com
Defining analytics:
a conceptual framework Analytics’ rapid emergence a decade ago created a great deal of corporate interest, as well as confusion regarding its meaning. By Robert Rose 34 | ORMS Today
|
June 2016
Arguably, the term “analytics,” as it is widely used today, was introduced in a research report “Competing on Analytics” [1] by Tom Davenport et al. in May 2005, and its emergence into public view coincided with the introduction of Google Analytics on Nov. 14, 2005. As can be seen in the Google Trends chart shown in Figure 1, in November 2005 searches for the term “analytics” jumped almost 500 percent. Davenport’s subsequent Harvard Business Review paper [2] and book [3] with the same title, as well as analytics-oriented marketing programs by IBM and other companies, further raised consciousness of the word and contributed to the subsequent dramatic growth in Google searches for the term analytics. The dramatic growth in the use of the term analytics has been accompanied by a proliferation in the way analytics ormstoday.informs.org
is used, and phrases such as “text analytics” and “healthcare analytics” have become common. Unfortunately, in addition to the great interest and excitement sur rounding analytics, there is a corresponding amount of confusion a n d u n c e r t a i n t y re g a rd i n g i t s meaning. Perhaps the best example of this is a statement from the beginning of a March 2011 article in Analytics magazine [4]: “It’s not likely that we’ll ever ar r ive at a conclusive definition of analytics,” a Figure 1: Google Trends chart shows the rapid rise in “analytics” searches. reference to surveys of INFORMS members who at the time expressed widely divergent views on the relationship analytics that are cited always examples of methods between operations research and analytics. from disciplines such as statistics, computer science, According to the survey: operations research, economics or industrial • A significant number of respondents engineering? believed that operations research is part of Analytics is sometimes represented as a conanalytics. vergence of the quantitative decision sciences. If • A significant number of respondents this is the case, it would represent a reversal of a believed that analytics is part of operations trend in human history, lasting for thousands of research. years, toward specialization. What caused such a • Some respondents believed that operations reversal, and why did it occur suddenly? More research and analytics are the same thing. importantly, since there is no new high-level • Some respondents believed that operations unifying theory associated with analytics, such as research and analytics are completely string theory in physics, how could individuals different disciplines. acquire the knowledge and master the methods of at least five or six separate disciplines? This lack of consensus regarding the To overcome the uncertainty surrounding anrelationship of analytics to other disciplines alytics, we need an overarching conception of anis not limited to INFORMS members. In a alytics that answers the preceding questions while September 2014 article in the European Journal remaining consistent with how the term is used. of Operational Research [5], the authors state: “One Moreover, since the terms “descriptive analytics,” possible reason for the discrepancy between “predictive analytics” and “prescriptive analytics” are the perceived opportunity analytics may offer often used interchangeably with data science, and to the OR/MS community and the amount since prescriptive analytics is often associated with of research in the area, as alluded to in above, operations research, a framework is needed to relate may be the lack of any clear consensus about all of these terms. analytics’ precise definition, and how it differs Analytics, Analytics, Analytics from related concepts.” The first step in gaining such an understanding is Unanswered Questions the recognition that the term “analytics” is used in The uncertainty surrounding analytics has led to at least three different ways, and therefore, requires several vague conceptions regarding the term and three separate definitions: many unanswered questions. Analytics is often 1. Analytics is used as a synonym for statistics referred to as an emerging field or an emerging or metrics. Examples are website analytics discipline. If analytics is an emerging discipline, (how many views or clicks) or scoring where did it come from, and how could it have analytics (number of points scored per 100 emerged suddenly? Why is there no unique possessions). research associated with analytics – as distinct from 2. Analytics is used as a synonym for data science. research associated with statistics, computer science Examples are data analytics and predictive or operations research? Why are the examples of analytics.
If analytics is
represented as a
convergence of the
quantitative decision sciences, it would
represent a reversal of a
trend toward specialization.
June 2016
|
ORMS Today
| 35
Defining Analytics
Basing our
3. Analytics is used in a very general way to represent a quantitative approach to organizational decision-making. This is Davenport’s “Competing on Analytics” usage.
understanding of analytics
on the way the term is being used
will allow us to think
more clearly about analytics, but there is still
a problem.
Attempting to cover all three of these usages with a single definition has led to much of the confusion surrounding analytics. The first usage refers to a type of Figure 2: Visualizing separate disciplines of science that have common measurement or counting. elements. The second usage refers to the processing of large amounts of data with the fact that disciplines such as statistics, computer advanced software technologies and sophisticated science, operations research, industrial engineering statistical and computer science techniques. The and economics continue to exist? third usage refers to a management philosophy that emphasizes a quantitative approach to decision- An Analogy making. To explain this apparent paradox, an analogy Basing our understanding of analytics on the will be helpful. If you see the words chemistry, way the term is being used will allow us to think biology, science, physics, geology and astronomy, more clearly about analytics, but there is still a are you confused about their meanings or their problem: embedded in the third usage is an aprelationship to each other? I think not. Further, I parent paradox. suspect you will visualize something similar to the Among analytics thought leaders there is one diagram shown in Figure 2. area in which there is agreement – analytics is reThe word “science” conceptually groups lated to many different disciplines: together the natural sciences. Although physics, • Davenport, Cohn and Jackson, in the chemistry, astronomy, geology and biology are all previously mentioned May 2005 research separate disciplines that use different methodologies report “Competing on Analytics” [1] offer and require the mastery of large amounts of statistics, operations research, industrial discipline-specific knowledge, they do have elements engineering, econometrics and mathematical in common. Each of them uses the scientific modeling as examples of analytics. method, mathematical modeling and peer-reviewed • Rahul Saxena, coauthor of the December 2012 book “Business Analytics,” on slide No. 5 of a SlideShare presentation [6], lists 14 disciplines as being antecedents of analytics. The list includes business intelligence, computer science, statistics, operations research, industrial engineering, and finance planning and analysis. If these author s are correct, how can we explain
36 | ORMS Today
|
June 2016
Figure 3: Visualizing separate disciplines of analytics that have common elements. ormstoday.informs.org
research. It is therefore appropriate to group them together; it is also quite useful to be able to refer to the natural sciences collectively. We can talk about the state of science education or discuss whether or not we should increase our investment in scientific research. Paradox Explained For the same reason that it is useful to have a term that collectively represents the natural sciences, it is useful to have a term that collectively represents the quantitative decision sciences. And this is exactly the role played by the term analytics in its broadest usage: It conceptually groups together the quantitative decision sciences (see Figure 3). Statistics, data science, industrial engineering, operations research and computer science are separate disciplines that use different methodologies and require the mastery of large amounts of discipline-specific knowledge. As in the case of the natural sciences, they do have elements in common such as the scientific method, mathAnalytics includes many disciplines, while at the same time, those disciplines continue to exist. ematical modeling and peer-reImage © Sergey Nivens | 123rf.com viewed research. When viewed as a conceptual grouping of the quantitative decision research. To create such a framework, we need one sciences, the term analytics, in its broadest usage, more important concept. allows us to make statements such as: “We will compete on analytics.” We can meaningfully refer Problem Centricity collectively to different quantitative decision sciIn a Dec. 17, 2014, INFORMS podcast [7], Glenn ences, possibly in different departments or differWegryn observes that analytics is divided into two ent geographical locations, in the same way that distinct camps. He notes that they tend to come we might collectively refer to separate scientific from different organizational backgrounds, and he research projects. describes them in the following way: Viewing analytics in this way explains the • data centric – use data to find interesting paradox of how analytics can somehow include insights and information to predict or anticipate many disciplines, while at the same time, those what might happen; and disciplines continue to exist. This understanding • decision centric – understand the business allows us to define analytics and explain its problem, then determine the specific relationship to the quantitative decision sciences. methodologies and information needed to solve Also, it is now possible to explain the sudden the specific problem. emergence of analytics: In the age of the Internet, new concepts can suddenly emerge and go viral. As is clear from its description, the decision-cenWe now have an overarching conception of tric category could also be named problem centric, what analytics is, but we do not yet have a frameand to make it clear that it encompasses systems and work that can relate the broadest usage of analytprocesses, that is how I will refer to it. Since analytics, ics to the various “flavors” of analytics (descriptive, in its broadest usage, conceptually groups together predictive, prescriptive), data science and operations the quantitative decision sciences, the data-centric June 2016
|
ORMS Today
| 37
Defining Analytics
Analytics
and problem-centric classification can also be applied to the quantitative decision conceptually sciences. Speaking about data science, Anthony Goldbloom (founder of Kaggle) said, “You want to extract all the signal that’s possible out the of a dataset” [8]. In a recent interview in OR/MS Today [9], Professor Edward Kaplan (president of INFORMS) said, “Operations researchers think in terms of problems” and “operations research is Figure 4: Conceptual framework of analytics. it represents the scientific study of operations.” These descriptions suggest that data science technologies and sophisticated statistical and (data centric) and operations research (problem cencomputer science techniques. but is not itself tric) fit nicely into the preceding classification. • Prescriptive analytics is the domain where data-centric and problem-centric paradigms A Conceptual Framework intersect, i.e., problems that require scientific Keeping the above in mind, and remembering that study and mathematical modeling, and the one of the usages of the term analytics is as a synprocessing of large amounts of data with onym for data science, a conceptual framework can advanced computer science and statistical now be constructed that relates analytics, descriptive techniques. analytics, predictive analytics, prescriptive analytics, • Many problems that require scientific study and data science and operations research (See Figure 4). modeling analysis do not require the processing Several aspects of the diagram shown in Figure 4 of large amounts of data and are represented by should be noted: the category prescriptive quantitative analysis. • Since the term analytics is used in multiple Summary ways, there is no conflict caused by its use above the diagram (broadly referring to quantitative The uncertainty surrounding analytics can be elimidecision-making) and within the diagram (a nated by keeping the following points in mind: synonym for data science). • Analytics is used in three different ways and • Data science, and its two-word analytics therefore requires three definitions. synonyms, refer to the processing of large • In its broadest usage, analytics conceptually amounts of data with advanced software groups together the quantitative decision sciences; it represents disciplines, but is not itself NOTES & REFERENCES a discipline.Therefore, there is no research that is unique to analytics, and there are no methods 1. Davenport, T.H., Cohen, D., Jacobson, A., “Competing on Analytics,” Babson College Research Report, May 2005. that are unique to analytics. 2. Davenport, T.H., “Competing on Analytics,” Harvard Business Review, January 2006, pp. 99• Analytics emerged suddenly since it is a concept 107. that went viral. 3. Davenport, T.H., Harris, J.G., 2007, “Competing on Analytics: The New Science of Winning,” • Operations research is a problem-centric Harvard Business Review Press. discipline; data science is a data-centric 4. Boyd, E.A., “What is analytics?,” 2011, Analytics magazine, March/April 2011, pp. 7-8. discipline. Prescriptive analytics is where 5. Mortenson, M.J., Doherty, N.F., Robinson, S, 2015, “Operational research from Taylorism to data-centric and problem-centric paradigms Terabytes: A research agenda for the analytics age,” European Journal of Operational Research, Vol. 241, Issue No. 3, March 2015, pp. 583-595. intersect. ORMS
groups together
quantitative decision sciences; disciplines,
a discipline.
6. Rahul Saxena, “Building an Analytics CoE (Center of Excellence),” SlideShare presentation, July 20, 2013. 7. Glenn Wegryn, “Top 5 Analytics Trends for 2015,” INFORMS Podcast, Dec. 17, 2014. 8. Anthony Goldbloom, “Lessons from 2MM machine learning models,” YouTube video, Dec. 11, 2015. 9. Horner, P., 2015, “Meet the ‘member in chief,’ ” OR/MS Today, Vol. 42, No. 6, pp. 36-41, December 2015.
38 | ORMS Today
|
June 2016
Robert Rose (robertl.rose@verizon.net) has been successfully using operations research to solve business problems for many years. A longtime member of INFORMS, he has formed a company, Optimal Decisions LLC, where he is currently developing O.R.-based decision support tools in the areas of employee scheduling and forward buying optimization.
ormstoday.informs.org
®
CERTIFIED ANALYTICS PROFESSIONAL Analyze What CAP Can Do For You ®
www.certifiedanalytics.org
Sports analytics taxonomy, V1.0
By Gary Cokins, Walt DeGrange, Stephen Chambal and Russell Walker
Image © Steven Gaertner | 123rf.com
Classification techniques provide an important first step in the serious study of the fast-growing field of sports analytics.
The application of analytics to sports has been going on for decades, long before computers came along. Major League Baseball (MLB) fans, for example, have been “keeping a scorecard” (with pencil and paper) at games for more than a hundred years, dutifully recording balls, strikes, outs, runs, hits, errors and other data. Today, in the era of advanced analytics and big data, MLB teams can employ multiple high-speed cameras and in some cases sensors to capture not only every minute detail of the game (speed and location of a pitch, direction and distance of a hit), but to correlate 40 | ORMS Today
|
June 2016
ormstoday.informs.org
Figure 1: Version 1.0 of the sports taxonomy. Source: Ben Grannan
every action and player movement with associated conditions such as the game situation (two outs, runner in scoring position, bottom of the ninth, one run behind, playing a night game on the road against a left-handed pitcher). The same is true for basketball, soccer and other professional sports teams and leagues. Why Collect and Analyze Sports Data? The explosive growth of advanced analytics in the professional sports world is mindboggling, prompting some to ask: What is the use of such minutia? It is a valid question, but when answers to this question begin to surface, one begins to observe that portions of the collected data support better decision-making. But that introduces another question: â&#x20AC;&#x153;What types of decisions?â&#x20AC;? For an MLB baseball manager, the decision might be as simple as whether a batter should bunt or hit away given the game situation. For an MLB general manager, the decision might be, based on data-driven player evaluations, how much should the team offer to
pay that player during salary negotiations or should the team consider trading that player to another team, and what player should the team try to get in return? The opportunities to apply analytics in sports extend beyond player evaluations and in-game coaching decisions and strategy. Examples include management of the sports franchise and venue, such as how much food to order for vendor concessions at the stadium, and how to best price tickets and corporate sponsorships. Additionally, analytics has given rise to a large investment in player biometrics (in order to monitor playersâ&#x20AC;&#x2122; health, including the severity of head concussions or heartbeat during extreme heat). Given this, sports teams deploy analytics at many levels (for players and during the execution of a specific play) and in many realms (from player performance to revenue management). How can anyone make sense of all the possible applications of analytics in sports? How many categories of sports analytics are there? Which type of statistical method is appropriate for each category? June 2016
|
ORMS Today
| 41
Sports Analytics With a
What level of a particular type of statistical method is practiced in sports today? Where are the opportunities to apply analytics in sports going forward?
coding scheme, articles and
other content can be
crossreferenced to organize
a “body of knowledge” of
sports analytics to aid
researchers.
Sports Analytics Taxonomy How do we answer these types of questions? The answer is to begin by creating a sports analytics taxonomy. A taxonomy is a technique for classifications; it’s typically hierarchical with a tree-branches-leaves structure. Examples are found in biology (plant and animal species), chemistry (basic elements and compounds), astronomy (types of galaxies and stars) and business organizations (types of industries and services). The taxonomy allows for a useful organization of the field. In this article we propose a version 1.0 of a sports analytics taxonomy. Its purpose will be not only to define the classifications of uses of analytics in sports, but also to identify applicable statistical methods for each classification, thereby promoting best in class research and application of analytics in sports.With an associated coding scheme, articles and other digital content can be cross-referenced to organize a “body of knowledge” of sports analytics to aid researchers. Going further, the taxonomy can be used to assess the current stage of maturity of various analytical methods in practice. One purpose for this assessment is to identify what opportunities exist for further application of analytics or to refine existing ones in sports. Figure 1 depicts the authors’ version 1.0 of the sports taxonomy. Rather than trying to describe every detailed branch and leaf of the taxonomy, this article is intended to describe the reasoning for the major “tree branches” of the taxonomy and then provide a few examples of some of the “tree branches.”The eight major tree branches depicted can be broadly grouped into three super-branches as follows: 1. The first three branches are team, individual and league sports.What these three branches have in common is “sports,” implying competition. 2. The fourth branch is recreational with a personal focus that is oriented toward an individual’s health and performance. 3. The last four branches involve the quest to conquer uncertainty.They are fantasy sports, sports betting, games of chance and professional online gaming. We considered alternative branches such as differentiating professional sports from amateur sports. Upon further examination, it became apparent that there would be substantial similarities between the two, creating redundant “branches and leaves.” On the other hand, team and individual sports trunks reduce redundancies since each
42 | ORMS Today
|
June 2016
trunk involves different types of decision-making. Consider this in regards to super-branch No. 1: • Team sports: Examples are familiar, including soccer, baseball and hockey. Some of the teamrelated minor branches are winning strategies; recruiting and scouting players; business operations, including stadium management and ticket pricing; and player evaluation for salary negotiations. Note that each of these mostly apply regardless of the players’ age level (youth, high school, college or professional). • Individual sports: Examples are golf and tennis. Some of the individual-related minor branches are body conditioning, biometric physio monitoring and behavioral modeling. • League sports management: The third branch of the “sports” grouping superbranch No. 1 involves the coordination of teams (but in some cases individuals) and has business management aspects to it such as scheduling, revenues,TV licensing and brand licensing. There will always be some similarities in each of the first two minor branches, for example strategy to win and body conditioning. High-Tech Revolution Sensor technology and strategically placed highspeed cameras – briefly mentioned at the beginning of this article – have clearly stimulated greater interest in the field of sports analytics for one obvious reason: they provide more data! One example, again from MLB, is Statcast, a technology that provides many cameras and radars located with different view angles in baseball stadiums. Data is collected at 30 frames per second for every player on the field, both defensive and “at bat,” as well as of the baseball. With this technology everything that is happening on the playing field is collected. Therefore anything can be measured. The acceleration of players running the bases or reacting to an outfield fly ball can be measured.The velocity, arc and accuracy of a shortstop’s throw can be measured. An outfielder’s path to make a play on a fly ball can be computed for efficiency. With this type of data, fan arguments about “who is the best player in a position?” will be bolstered with the facts. But a team’s management might also be able to better protect its assets. As an example, if a pitcher’s throwing arm delivery starts deviating from its normal pattern, it could provide an alert warning that the pitcher strained a leg muscle, and that further throwing could lead to a severe season-ending surgery. What will teams and leagues do with all that data? That is both the fun and the conundrum of ormstoday.informs.org
Image © Eric Broder Van Dyke | 123rf.com
Sensor technology and strategically placed high-speed cameras capture enormous amounts of information, whether it is professional basketball (above), baseball or other sports.
sports analytics. The breadth of uses of sensor technologies in sports is unimaginable. Sensors and high-tech cameras focused on the playing field bring up a controversial topic worth mentioning. Thanks to instant replay and sensors, television viewers already see precisely whether a tennis serve is in or out, whether a baseball pitch is a ball or a strike, or whether a football receiver got both of his feet down in bounds on a catch near the sideline. The same technology that is collecting data can also be used for “rules enforcement.” Will such technology eliminate the need for umpires, referees and other game officials? Wouldn’t we miss a baseball umpire raising his thumb and yelling “yourrrrrree out!” when a base runner slides into home plate just as the catcher applies the tag? Before closing, the recreational and gaming/betting super-branches of the taxonomy tree of sports analytics deserve some mention.The recreational super-branch applies analytics with little or no emphasis on winning to avoid losing. Here, the purpose is to support personal health and nutrition.Think digitized treadmills, Fitbit and other wearable devices that not only monitor one’s food intake and calorie burn, but also provide periodic analysis.The gaming and betting super-branch applies analytics to fantasy league player drafting, betting point spreads, cheating detection and many more.
Where Do We Go from Here? The wide breadth of applying sports analytics is apparent and possibly overwhelming, but that is a good thing. It means the opportunities for researchers and analysts are many. What this article is intended to convey is the need to organize the “body of knowledge” – developing an initial structured taxonomy – for sports analytics to enable more efficient research and application of the tsunami of sports data that is approaching. ORMS Gary Cokins (gcokins@garycokins.com) retired from SAS in 2013. He is founder of Analytics Based Performance Management LLC (www.garycokins.com). Walt DeGrange (wdegrange@canallc.com) is a principal operations research analyst at CANA Advisors (www. canallc.com) and the chairperson for the INFORMS SpORts Section.
The wide breadth of
applying sports analytics is apparent and
possibly overwhelming, but that
is a good thing.
Stephen Chambal (stephen.chambal@theperducogroup. com) is the CEO of The Perduco Group (www. theperducogroup.com), a high-end data analytics company working in defense, healthcare and sports industries. Russell Walker (russell-walker@kellogg.northwestern.edu) is clinical associate professor of managerial economics and decision sciences at the Kellogg School of Management of Northwestern University. He can be reached at http://www. russellwalkerphd.com/. All four co-authors are members of the INFORMS Section on OR in Sports (SpORts).
June 2016
|
ORMS Today
| 43
S
FORECASTING SOFTWARE SURVEY
New tools, new capabilities, new trends: Survey of 26 software packages from 19 vendors. Image © seewhatmitchsee | 123rf.com
By Chris Fry and Vijay Mehrotra
Forecasting 2016 Welcome to the 2016 OR/MS Today Forecasting Software Survey, where we take stock of the latest technologies and trends in forecasting affecting both vendors and users.The data used to develop this survey includes responses from 19 vendors for 26 software packages that span a range of capabilities and price points, with the results from this survey summarized below. As part of the rapid proliferation of business analytics, forecasting models are an increasingly important part of the management landscape. More and more managers and executives rely on sophisticated forecasting methods, not only for planning purposes but also as a foundation for performance analysis, process improvement and optimization. Open source software continues to be a major factor driving the growth of analytics, offering a unique combination of flexibility, power and low cost. In particular, many leading firms are using tools like R and Python to develop solutions for forecasting and predictive analytics that are customized for their business problems, tightly coupled with their data architectures, and integrated directly into other existing systems. Given all of this, the market for forecasting software vendors is dynamic, featuring many different types of innovations. Here are some of the major trends that we are seeing in the marketplace. Top Forecasting Trends for 2016 No. 1: integration. Corporations are seeking not only to generate forecasts, but also to integrate those forecasts into planning, optimization and reporting systems. As one survey respondent, ForecastPro, stated,“More and more business users are calling for 44 | ORMS Today
|
June 2016
comprehensive forecasting solutions that in addition to simply generating statistical forecasts can be used as the backbone of ongoing corporate processes such as sales and operations planning (S&OP), demand planning and supply chain optimization.” Many vendors appear to be addressing this demand. No. 2: automation. Forecasters are seeking features such as automated model selection, automated alerts, and automated graphics and reporting. As data velocity and complexity are growing, there appears to be an increasing willingness to entrust model design to the software, especially when many forecasts are being generated simultaneously. In response, nearly all of the vendors surveyed report having some type of automatic or semi-automatic forecasting capabilities. No. 3: visualization. High-quality visualizations are fast becoming part of the “table stakes” for forecasting software packages. In addition to standard statistical output, many of today’s tools offer a range of visuals, including such features as box plots, normal probability plots, histograms, ANOVA Pareto charts, decomposition charts and automated statistically generated range forecast plots. No. 4: virtualization. Several vendors have begun to move their offerings to the cloud, offering virtual hosted forecasting ormstoday.informs.org
S tools. Statgraphics, XLMiner.com, Roadmap GPS and PEERForecaster/PEERPlanner all mentioned their cloud offerings specifically in their survey responses. As cloud computing continues to grow, we foresee this trend continuing. No. 5: forecast quality measurement. Forecasters are continually pressured to improve the quality of their forecasts, or defend why their forecasts are as good as they can get.Vendors are addressing this need with such solutions as automated ANOVA analysis, automated naïve forecast generation and automated forecast value added analysis. No. 6: capabilities enhancement. Forecasting software vendors are giving increased attention to “hard” forecasting problems such as new product forecasting and forecasting of intermittent demand, while also providing ways to integrate additional machine-learning techniques into their forecasting suites. Some vendors are offering the ability to create automated ensemble forecast models, built by combining multiple forecasts generated using different techniques in order to improve overall forecast accuracy. Case Study: Forecasting Rainfall in California Figure 1: Monthly rainfall for a sample collection station in California, along with the corresponding El Niño indices. To illustrate some of these trends, we attempted to forecast monthly rainfall levels in our home state of California, which has been suffering from a drought for the past few years. After collecting and prepping the data, we utilized the Expert We utilized publically available data from the National Oce- Modeler to generate an initial set of forecasts. SPSS created a sepanic and Atmospheric Administration (NOAA) [1], collecting arate independent model for each weather station, with total run monthly precipitation data (in inches) from January 2006 through time less than one minute. SPSS modeled the majority of stations March 2016 for 194 observation stations in California. From this, (177 out of 194) using simple seasonal exponential smoothing. For we used the first 10 years of data as a training set and then at- the remainder of the stations, it used a combination of ARIMA, tempted to forecast the monthly rainfall at each station for the final seasonal ARIMA and ARIMA with transfer function models three months (i.e., January-March 2016).We also included one ex- (to incorporate the El Niño index).The model structures for the ogenous variable, NOAA’s El Niño Index [2] (the El Niño climate remaining models varied widely, from simple “flat line” models oscillation is a well-known factor in predicting California rainfall, (when the software could not find a suitable pattern) to complex as changing temperatures in the Pacific Ocean affect rain patterns multi-term ARIMA models such as ARIMA (0,0,3) (1,1,0) with on the West Coast). 13-month delayed seasonally differenced external regressor effect Figure 1 shows the rainfall history for a sample collection or ARIMA (0,0,11) (0,0,0). In addition to the forecast values, the station, along with the El Niño Index history. The rainfall fol- software also produced goodness of fit statistics for individual and lows a somewhat seasonal pattern at this station, and appears aggregate models, along with parameter estimates and plots. to show some (slightly lagged) relationship with the El Niño Visually, the fitted time series looked reasonable when comindex. Note the high El Niño activity toward the end of 2015 pared to the actuals. Figure 2 shows an example fit, along with a and corresponding heavy rainfall level in January 2016. three-month projection, for a sample collection station. We selected IBM’s SPSS Statistics forecasting package to conWe then spent additional time to review and refine the models. duct our analysis.We had not used SPSS previously, so we felt that We noted that the software’s initial fitted model output included negthis choice allowed us to work without any preconceived expecta- ative rainfall estimates in some months, so we adjusted these to zero in tions on how the software would perform, and would also enable our analysis.We also rejected some of the forecasts in favor of models us to convey a realistic experience in terms of the learning curve that we felt were either more intuitive or would better capture the required to use the product. IBM offers a generous, fully functional seasonal behavior and El Niño effect. Shown in Figure 3 are two “betwo-week free trial for the SPSS product, which made it easy for fore and after” examples in which we replaced an automatically-genus to download and start working with the product right away.The erated seasonal exponential smoothing model with a seasonal ARIMA package includes a fully automatic forecasting module, called the model including transfer effects for the El Niño external regressor.The Expert Modeler, which optimizes model and parameter selection new models seemed to fit the seasonal peaks better than the smoothacross a suite of exponential smoothing and ARIMA models. ing models did, and also projected higher 2016 rainfall in response to June 2016
|
ORMS Today
| 45
S
FORECASTING SOFTWARE SURVEY
In all, we felt that despite our limited data set, the automated procedure gave us a very quick firstpass at the analysis that seemed quite reasonable, and may likely be satisfactory in many contexts. The experience reconfirmed to us the power of modern automatic forecasting tools, and also reminded us of the value that comes from coupling that power with our time and attention as analysts to continually seek out improvements in our models. Happy forecasting to you in 2016! About the Survey For this year’s forecasting software survey, as in the past, OR/MS Today attempted to include as many forecasting products as possible. We contacted all prior survey participants, as well as any new vendors that we were Figure 2: Plot showing fitted time series and forecasted rainfall compared to able to identify.We asked each respondent to complete actual rainfall for a sample collection station an online questionnaire covering a comprehensive list of questions spanning features and capabilities, recent the high El Niño Index. Figure 3 shows example model fit and fore- enhancements, licensing and fees, technical support and other areas. cast plots for two stations, comparing the SPSS Expert Forecast output OR/MS Today followed up with each vendor to help ensure that with that of alternate models that we selected instead. we obtained as many responses as possible.Vendors not included Software Survey, continued on p.52
Figure 3: Sample model fit and forecast plots comparing SPSS’ automatic forecast output with potential alternate model structures. For the first example (San Jose), the auto-generated model predicted 6.97 inches of rainfall for Q1 2016, while the alternate model predicted 7.86 inches. Actual rainfall was 8.36 inches. For the second example, projected Q1 2016 rainfall was 5.61 inches for the automatic model and 7.39 inches for the alternate model. Actual rainfall was 7.22 inches.
46 | ORMS Today
|
June 2016
ormstoday.informs.org
What’s Your StORy? Chanaka Edirisinghe Kay & Jackson Tai '72 Chair Professor at Lally School of Management of Rensselaer Polytechnic Institute, NY Member of INFORMS since the late 1980s What prompted you to enter this field? Why? I can honestly say that my basic foundation for this field was formed during my undergraduate days in mechanical engineering in Sri Lanka when I first took a course in operations research, in which I studied the simplex method developed by Professor George Dantzig. Later, during my master’s work in industrial engineering, I fell in love with the field of optimization and stochastics, partly due to my first real OR practical application in water reservoir design and operation for the energy authority in Thailand, which led to furthering my interests in this field at UBC where I completed my doctoral work in stochastic optimization. Professor Dantzig remains my role model to this day. What INFORMS member benefit do you find the most useful? First and foremost, networking opportunities for junior researchers are tremendous. As a young researcher, I found it extremely rewarding. Second, INFORMS is an exciting place to sharpen leadership skills within ones’ own passionate domain of OR/MS interest. These are provided through many societies and subdivisions. A wealth of free information is also available online for INFORMS members via TutORials, publications, and other resources. What are you most excited about as chair of the 2016 INFORMS Annual Meeting? It is exciting to have INFORMS return to Nashville after 25 years – the last was in 1991. I learned a long time ago that success does not come by just working hard, but enjoying what you do is part and parcel. The most exciting part of this Annual Meeting to me is it is “INFORMS” (working hard) in “Nashville” (enjoyable). While we work hard to “fine tune decisions in the Music City,” I hope that each and every attendee will enjoy the superb cuisine and the variety of entertainment options that Nashville has to offer. So, I feel fortunate to be chairing this great meeting. I am also supported by a terrific core organizing committee of 14 people, and a great INFORMS staff. What is something you learned in the last week? Life is too short, for I narrowly escaped a road accident in San Francisco. On a more serious note, it reminds me that it is not worth doing things we don’t enjoy and feel proud of.
More questions for Chanaka ? Ask him in the Open Forum on INFORMS Connect!
http://connect.informs.org
Analytica
Windows NT, XP, Vista, 7, 8, 10
Autobox
Windows 10, R, Unix, AIX, Solaris, Linux, MAC in an emulator Windows 7, 8, 8.1, 10
Lumina Decision Systems, Inc.
Automatic Forecasting Systems, Inc.
Crystal Ball Predictor Oracle Corporation
32-bit, – y y – 64-bit Windows
–
Windows XP, Vista, 7, 8, 8.1, 10
Syncopation Software
Graphics Types
Techniques Available I. Exponential Smoothing
y y
–
CSV
64-bit works on all
y y y y 10,000 y –
–
–
32- or 64-bit
y y y –
–
y –
–
32- or – – y – 64-bit Windows
–
y y
–
CSV, ODBC
y y Bitmap, y y y – – – y Tornado Windows diagrams, Metafile risk (PowerPoint profiles, optimized) bubble charts, etc.
– – – – – –
(must match Excel bitness)
DPL
Export
Brown’s Winter’ , Holt’s Two-Pa s Three -Param rameter, eter Me Adaptiv thods e Respo nse Rate Harriso n’s Harm onic Sm Dampe oothing d Trend Program finds Op timal Sm Other fe oothing a statistic tures? (e.g., use [AIC, BIC r c minimiz , root M an specify e S model b d, software cho E, etc.] to be ased on oses ap propriate characte o ristics, e ther [specify] tc.
Import Read Excel 2013, 2010 or older?
2013 an d 2010 2007 an d 2003 If not, w made? hat modificatio (i version, .e., user must ns must be sa insert a dditiona ve as earlier l inform Addition ation, etc a l .) im port (e.g., sa v, mtw, file formats a V9, etc.) vailable ? Exports Output to Exce Exports l F il e? Session Output? Exports graphic s? (Spe cify form ats) Time-se ries Plo t Scatter Diagram Forecas t Plot Autocorr elation, Partial Autocorr elation, Functio Residua n l Plot Bar, Pie , other B usiness Chart Others (Please specify)
32-bit, 6 4-bit, Platform (e.g.,
(i.e. Win d Mounta ows 8.1, 7, Ma in Unix, etc Lion, Lion, Sn c OS [Maverick ow Leop , .) ard],
Software Product Listing
Software Data Entry Capabilities
etc.) Automa tic Semi-A utomati c Manual Max # o f Observ ations in Time-se If yes, h ries? ow man y?
Operating System(s)
y – EMF, y y y y – y y Tornado PNG, chart, JPEG, waterfall BMP, chart, heat TIFF, and map more
y – – – – Libraries available with most smoothing methods
y y PNG and JPG
– – – – – Model chosen based on tests of Gaussian Assumptions
Any file y y readable by Excel
(works on both)
–
y y y y y y – Actual and Cleansed of Outliers
y – – y y – –
–
y – – y y –
Forecast Pro TRAC
All Windowsbased operating systems
Ships y y y – with both 64- and 32-bit versions
–
y y
–
CSV, ODBC
y y Output y – y – – – y Year-overcan be Year and saved Other to Excel Customized and to Graphs and PNG file Plots
y – – y y Croston’s Intermittent Model, Seasonal Simplification Model, Discrete Data Models
Forecast Pro Unlimited
All Windowsbased operating systems
Ships y y y – with both 64- and 32-bit versions
–
y y
–
CSV, ODBC
y y Output y – y – – – y Year-overcan be Year and saved Other to Excel Customized and to Graphs and PNG file Plots
y – – y y Croston’s Intermittent Model, Seasonal Simplification Model, Discrete Data Models
Forecast Pro XE
All Windowsbased operating systems
Ships y y y – with both 64- and 32-bit versions
–
y y
–
CSV, ODBC
y y Output y – y y y y y Year-over-Year, can be Cumulative saved and Other to Excel Customized and to Graphs and PNG file Plots
y – – y y Croston’s Intermittent Model, Seasonal Simplification Model, Discrete Data Models
IBM SPSS Forecasting
Windows, Mac, Linux
32-bit, y y y – 64-bit, IBM z Systems, POWER
–
y y
–
SPSS y y BMP, y y y y y y y data file, TIF, PNG, SAV, SAS JPG, data file, EMF stata data file, etc.
–
y – – y – –
i-Data
Most recent Windows, Unix and Linux
Both
y y y –
–
y y
–
Several y y GIF, others JPEG, includPNG ing XML, SQL, CSV
–
y y y y y –
Minitab Statistical Software
Windows 7, 8, 8.1, 10
Both 32-bit and 64-bit
– y y –
–
y y
–
MTW, y y JPG, y y y y y y y Histogram, XML, PNG, TIF, dotplot, TXT, CSV, BMP, GIF, boxplot, DAT, EMF interval dataplots, 3D bases plots, etc.
y – – – y –
NCSS
Windows 7,8,8.1, 10
32-bit or y y y – 64-bit
–
y y
–
over 30 y y WMF, y y y y y y y Many, formats EMF, many, JPG, and more more
y – – y y –
Business Forecast Systems, Inc.
Business Forecast Systems, Inc.
Business Forecast Systems, Inc.
IBM
MJC2
Minitab, Inc.
NCSS
48 | ORMS Today
|
June 2016
y y y y y y y
ormstoday.informs.org
Techniques Available
Documentation
y – y
y
50
– – –
y
149
y y –
y
No limit
– – –
–
–
y – –
y
y – –
–
No limit
Forward – – – & Iterative Stepwise
–
New Features
Comments
S
(Since June 2014)
– – – – y y y – –
–
– – – – – – – – –
–
CART
–
–
– y
–
y y All outliers are reported to provide BI, simulated forecasts, etc.
Can Use r Down load Tria l Versio n?
Educati on
Comme rcial
(e.g. multivariate techniques, etc.)
Stepwise y – y Logistic, – – y – – – y – – Probit
10,000 Transfer – y – Function models
Price
VI. Additional Capabilities
Automa tically fi nd the O Online H ptimal M elp and odel? Tutorial? Other F eatures (e.g., so ft software ware offers fore c statistic etc.) ally valid asting advice, ates the model,
V. Other Univariate Techniques
Other? (please specify)
IV. Regression Models
Maximu m # of In depend (If yes, h ent Vari ow man ables? y?) Maximu m # of O bservati (If yes, h ons? ow man y?) Other L inear R egressio (If yes, p n Mode lease sp ls? ecify) Nonline ar Regre ssion M Dynamic odels Regress ion Trend A nalysis Specify Trend M odels (i.e., exp onential, S-curve , Weibull, etc.) Spectra l Analys is Kalman Filter Fourier Analysis State S pace M odels Transfe r Functi on Mod Interven el tion Ana lysis Econom etric Mo d els SABL Neural Network s
Moving Average Method Classic s al Deco mpositio Census n Bureau Method s (i.e., X-12 -ARIMA) Program finds Op timal M odel
Time II. Series TimeDecomposition Series III. ARMIA Decomposition (Box-Jenkins)
Call
Approx. y New graphing options, half improved econometrics, commersensitivity analysis. cial price
Key advantages: Visual influence diagrams, Intelligent Arrays, Monte Carlo.
Call
20% y Simulated Forecasting - Identification of leads discount Resampling to generate and lags automatically
a distribution of for causals. forecasts for better C.I., Elasticities option.
$995 Class- y Automatic construction – y Full online room documenof random walk discounts tation includ. formulas in available statistical spreadsheet. Additional references & forecast adjustment examples options.
Has unique ability to perform risk analysis using Monte Carlo simulation with the forecast results
from $59
y Enhanced multiselect, display custom percentiles in reports, multiple objective functions, parallel endpoint recording.
Flexible forecasting and risk analysis with Monte Carlo and discrete tree methods.
–
–
– – –
– – – y – – y – – Influence – y diagrams, Monte Carlo simulation, sensitivity analysis
–
–
–
–
– – y Exponen- – – – – – – – – – Promotional y y tial, Modeling, S-Curve, Event Linear, Modeling, Bass Custom model, etc. Allocation
–
$8,995 Call for y Custom component quote modeling, new product forecasting models and substantial improvements throughout.
Forecast Pro TRAC is a comprehensive system offering substantial forecast management/ manipulation/tracking capabilities.
y
–
–
–
– – y Exponen- – – – – – – – – – Promotional y y tial, Modeling, S-Curve, Event Linear, Modeling, Bass Custom model, etc. Allocation
–
$4,995 Call for y Custom component quote modeling, new product forecasting models and substantial improvements throughout.
Forecast Pro Unlimited is an affordable, comprehensive, easyto-use forecasting package designed for large-scale forecasting.
y – –
y
100
–
–
y y y Exponen- – – – – – – – – – Promotional y y tial, Modeling, S-Curve, Event Linear, Modeling, Bass Custom model, etc. Allocation
–
$1,295 Call for y Custom component quote modeling, new product forecasting models and substantial improvements throughout.
An affordable, comprehensive, easyto-use forecasting package ideal for both business and academic users.
y y –
y
–
–
See Please y Temporal causal Website contact modeling; spatiolocal temporal modeling. IBM rep
–
y y y
–
–
–
y y –
y
y
–
Stepwise, y – y Linear, – – – – – – – – – Multivariate, – y Assistant $1,495 $29.99 y Assistant DOE, multiple Best quadratic, quality, guides you regression; regression, Subsets exponenANOVA, DOE, through GLM have automatic tial reliability, analysis & model selection, growth, nonparaalerts you of response optimizer, new S-Curve metrics violations. graphs, and more.
2,000
–
Ridge, y – y Expo, y – y – – – – – – robust, weibull, stepwise, logistic, twonegative stage binomial
y y –
y
Wide y – y variety of stepwise options, best subsets –
y – –
–
y y y – y y – – y
–
– y
–
–
y – – – – – – – –
–
y y
–
–
– y 100’s of other statistical procedures
from $745
POA
$479
POA
$379
– –
–
y Over 20 new statistical procedures.
June 2016
|
Minitab is the leading statistical software for quality improvement worldwide.
This is a full-featured statistical package.
ORMS Today
| 49
OMP Plus Forecasting
Windows 8.1, 7
OM Partners
Optimal Scientist
Windows/ DOS (any version); SCO UNIX OpenServer (any version)
Transpower Corporation
32-bit, 64-bit
y y y –
–
32-bit
– y – y 999
CPDF/Delphus
–
PEERPlanner Forecast Decision Support System
Windows 8.1, 7
32-bit y y – – Windows
–
Windows 7, 8, 10
SAS Forecast Server SAS
TXT, CSV, y y XLS, y – y – – y y Pivot chart XML, BMP, TSV, JPEG, PRN PDF, GIF
y – – y y User can specify statistic to be minimized (AIC, MAD, MAPE, MSE)
– –
–
Text – – Text y y y y y y y 3-D import export surface
– – – – y Determines optimal value of predictor variables and the resultant optimal regression eq.
y y Vanilla Excel VBA Add-in
y y Excel
y – y y y y – Prediction limits plotted with forecasts
y y WMF
y – y – – – –
All major All major y y y – PC, Unix, platforms and supported mainframe operating systems
–
y y
–
Can y y Any y y y y y y y Define import major graphics virtually image and access any file format them via format stored processes
y – – y y Multiplicative and additive seasonal smoothing; user can specify statistic
y y y –
–
y y
–
Can y y Any y y y y y y y SAS import major Enterprise virtually image Guide any file format provides format additional reporting
y – – y y Combination models; multiplicative and additive seasonal smoothing; etc.
one y y y – version that works on both
–
– y user must save as earlier version
64-bit
– –
CSV, TAB, DAT, MAD
y y WMF
Windows, Mac OS, Linux; any that supports a browser
Cloudbased
Any OS a Web-based application)
|
y y y y 20,000 y y
–
(accessed
CSV, SGD, TXT
y y
–
through web browser)
One y y y – version which works on both
June 2016
y y y y y y y Surface
gram, etc.
(including tablets)
(Vanguard is
y – y – – y y
CSV, DBF, y y BMP, y y y y y y y CrossDBT, DTA, GIF, JPG, correlation, JMP, MAT, PNG, TIF, high-lowMDB, WMF close, etc. periodo-
64-bit)
50 | ORMS Today
y y – y y Geneva Expert System
Micro- y BMP soft Access, SQLServer
(premium for
Vanguard Software
–
–
Windows Choice y y y y 10 y y XP, Vista, 7, of 32-bit Million 8, 10 and 64-bit or later versions
Vanguard Forecast Server
y – – y y –
y y
Statgraphics Centurion
Statpoint Technologies, Inc.
–
–
Windows 8.1, Linux
Statgraphics Stratus
y – – y y AIC, BIC optimization
32-/64-bit y – – –
SCA Statistical System
Statpoint Technologies, Inc.
–
ASCII – – SQL query, ASCII flat files, or cutpaste
Windows
Scientific Computing Associates Corp.
Brown’s Winter’ , Holt’s Two-Pa s Three -Param rameter, eter Me Adaptiv thods e Respo nse Rate Harriso n’s Harm onic Sm Dampe oothing d Trend Program finds Op timal Sm Other fe oothing a statistic tures? (e.g., use [AIC, BIC r c minimiz , root M an specify e S model b d, software cho E, etc.] to be ased on oses ap propriate characte o ristics, e ther [specify] tc.
–
SAS Forecasting for Desktop SAS
Techniques Available I. Exponential Smoothing
64-bit)
32-bit y y – y Windows
RoadMap Technologies, Inc.
Graphics Types
required for
Windows 8.1, 7
RoadMap Global Planning Solution
Export
y y
(DOSBox is
PEERForecaster Excel Add-in
Delphus, Inc.
Import Read Excel 2013, 2010 or older?
2013 an d 2010 2007 an d 2003 If not, w made? hat modificatio (i version, .e., user must ns must be sa insert a dditiona ve as earlier l inform Addition ation, etc a l .) im port (e.g., sa v, mtw, file formats a V9, etc.) vailable ? Exports Output to Exce Exports l F il e? Session Output? Exports graphic s? (Spe cify form ats) Time-se ries Plo t Scatter Diagram Forecas t Plot Autocorr elation, Partial Autocorr elation, Functio Residua n l Plot Bar, Pie , other B usiness Chart Others (Please specify)
32-bit, 6 4-bit, Platform (e.g.,
(i.e. Win d Mounta ows 8.1, 7, Ma in Unix, etc Lion, Lion, Sn c OS [Maverick ow Leop , .) ard],
Software Product Listing
Software Data Entry Capabilities
etc.) Automa tic Semi-A utomati c Manual Max # o f Observ ations in Time-se If yes, h ries? ow man y?
Operating System(s)
–
y y
–
CSV, TXT, y y PNG HTML, and others
y y y y y Automatic model selection, outlier adjustment, multiple constrained opt. methods, etc. y – – – y User can specify statistic to be minimized (AIC, HQC, SBIC, MSE, MAE, MAPE)
y y y y y y y High-lowclose plot, interactive smoothing, periodogram, etc.
y – – – y User specifies statistic to be minimized (AIC, HQC, SBIC, MSE, MAE, MAPE)
y y y y y y y Stackable charts. zoomable charts
y – – y y Software chooses the appropriate model and parameters, etc.
ormstoday.informs.org
Techniques Available
Documentation
–
– y –
–
– – – y y – – – –
– – –
y
999
999
–
y – –
–
– – – – – – – – – Multi-variate – y regression; performs D-optimal experiment design
y y –
y
–
–
–
– – –
–
– – – y – – – – –
y y –
–
–
–
–
y y Software points user to items with largest errors, requiring review –
–
–
$595 Win/ for Win/ DOS: DOS; $495 $995 for UNIX; other FREE
FREE
ERROR state (call for space formulation details) leading to asymmetrical prediction limits
(call for
y – Multiplicative
– State space models, ARIMAX model, higher performance of the system that automatically chooses the best model, ...
Part of an integrated supply chain planning suite, including S&OP, Forecast netting, Operational Planning.
– –
Optimal Scientist helps with model formation, experiment design, experiment analysis, and model validation.
y –
Best suited for learning about State Space Forecasting and benchmarking commercial software.
– Offered as rental service with customer database in the cloud.
–
details)
–
–
– – y State Space
– – – y – – – – –
–
– –
–
Cloud rental after setup fee for database
–
–
–
– – y Exponen- – – – – – – – – – tial, Logistic, S-Curve
–
y y
–
$10,000 Free – Integrated monthly/ per Public weekly forecasting and user Version planning; interface to (call for Tableau Visualization details) Software. (U.S. Only)
y y y
y
–
–
Stepwise, y y y Virtually y y y y y y y – – Neural PlatFree y y Regional and – Web interface international best any model networks form SAS provides standardized user groups; subsets, available available workflow, demand online demos, depen- University discussion & many through through SAS dent Edition classification & rulesforums, 24x7 (see others SAS/STAT Enterprise based segmentation, tech support; etc. website for available
y y y
y
–
y – –
y
–
y y y
y
1,024
y y –
y
100
y y –
y
No limits
S
Can Use r Down load Tria l Versio n?
Educati on
–
Comme rcial
–
–
Comments
(e.g. multivariate techniques, etc.)
y
y
New Features (Since June 2014)
y y –
– – –
Price
VI. Additional Capabilities
Automa tically fi nd the O Online H ptimal M elp and odel? Tutorial? Other F eatures (e.g., so ft software ware offers fore c statistic etc.) ally valid asting advice, ates the model,
V. Other Univariate Techniques
Other? (please specify)
IV. Regression Models
Maximu m # of In depend (If yes, h ent Vari ow man ables? y?) Maximu m # of O bservati (If yes, h ons? ow man y?) Other L inear R egressio (If yes, p n Mode lease sp ls? ecify) Nonline ar Regre ssion M Dynamic odels Regress ion Trend A nalysis Specify Trend M odels (i.e., exp onential, S-curve , Weibull, etc.) Spectra l Analys is Kalman Filter Fourier Analysis State S pace M odels Transfe r Functi on Mod Interven el tion Ana lysis Econom etric Mo d els SABL Neural Network s
Moving Average Method Classic s al Deco mpositio Census n Bureau Method s (i.e., X-12 -ARIMA) Program finds Op timal M odel
Time II. Series TimeDecomposition Series III. ARMIA Decomposition (Box-Jenkins)
Miner; etc.
details)
multistage modeling.
–
Stepwise, y y y Virtually y y y y y y y – – Neural y y Regional and Depends Contact – Performance international best any model networks on enhancements. SAS user groups; subsets, available available online demos, number details & many through through SAS of discussion forums, 24x7 others SAS/STAT Enterprise users tech support; etc. available Miner; etc.
–
Stepwise, y y y Logistic all possible, minmax, QR, ridge, etc.
y y y y y y y – y GAM, Projec- y y tion Pursuit, Alternating Condition Expectations, LOESS, etc.
–
$1,350 $895 – – and up and up
RoadMap GPS allows companies to combine weekly demand planning with monthly and quarterly financial forecasting. Fact sheets, whitepapers, webcasts, demos, customer success stories, and other information at www.sas. com/forecastserver Fact sheets, whitepapers, webcasts, demos, customer success stories, and other information at www.sas. com/forecasting. –
10 Best y – y Quadratic, y – y – – – – – y Multivariate y y Software $1,495 $30 for y Trading bands, buy-sell Over 250 procedures (32-bit) 6/mnths interprets Million subsets, exponenmodels signals, dynamic for predictive analytics significance $1,810 GLM, tial, visualizations, interface and data visualization. (64-bit) of statistical logistic, S-curve, to R. $790results etc. logistic, $950 probit (annual) 20,000 Stepwise, y – y Exponen- y – y – – – – – – Box-Cox, tial, Cochranequadratic, Orcutt S-Curve, logistic, probit No limits
–
–
– y Software interprets significance of statistical results
$550 per year
$30 for y Interactive Statlets for 6/mnths data visualization.
No-cost On y y y S-curve, y – y – y – y – – Croston’s y y Pop-up and options – Demand/supply hover-over request available Spreadintermittent planning, multi-echelon user guides through curve, demand model, inventory optimization, Acad. for features. Logistic, Addressable financial planning, Partneretc. Gompertz, mrkt. models, consensus forecasting, ship and others etc. IBP/S&OP workflows. Prog.
June 2016
|
Browser-based interface, accessible over the Internet.
For over 20 years, more than 3,400 companies in 68 countries have relied on Vanguardís forecasting and analytics solutions.
ORMS Today
| 51
Export
Graphics Types
XLMiner Analysis ToolPak
Any web browser, any operating system
– – y y Practical y y limit 10,000
–
Google y – Sheets version: native format
XLMiner Pro and Platform
Windows 32-bit or – y – y See Website: y y 10, 8, 7, 64-bit, www. Vista, user solver. Windows specifies com/ Server 2016, xlminer2012, 2008 limits
–
Imports y y PNG, y y y y y y y Display in from SQL JPEG, Power BI dataBMP, and bases, Apache etc. and Tableau Spark Excel Big Data charts clusters
XLMiner SDK Pro and Platform
Windows C++, C#, – y – y Prio y y 10, 8, 7, Java, R, version Vista, Python; has Windows 32-bit or limits, Server 2016, 64-bit, See Website 2012, 2008 (user specifies)
–
CSV, SQL y – databases, Apache Spark Big Data clusters
XLMiner.com
Any web browser, any operating system
–
Imports y y PNG, from SQL JPEG, dataBMP, bases, Apache etc. Spark Big Data clusters
Frontline Systems Inc.
Frontline Systems Inc.
Frontline Systems Inc.
Frontline Systems Inc.
S
FORECASTING
Excel Online, Google Sheets
–
– y – y See Website: y y www. solver. com/ xlminerlimits
|
–
–
– – – – – – – Line Fit plot in Google Sheets version
– – – – – Simple exponential smoothing, matches Excel’s Analysis ToolPak y – – – y –
– – – – – – – Development Kit, graphics supplied by calling application
y – – – y SDK highlevel objects allow many properties/ features to be chosen
y y – – – – y New, very powerful graphics options coming in Fall 2016
y – – – y –
Software Survey, continued from p.46
in this issue of OR/MS Today are invited to submit a completed online questionnaire (http://www.lionhrtpub.com/ancill/fssurvey. shtml), and their product will be added to the online version of the forecasting survey. The purpose of the survey is simply to inform the reader of what is available. The information in the survey comes directly from the vendors, and no attempt was made to verify or validate the information they gave us. Automation levels: Forecasting software varies when it comes to the degree to which the software can find the appropriate model and the optimal parameters of that model. For example, Winters’ method requires values for three smoothing constants, and Box-Jenkins models have to be specified with various parameters, such as ARIMA (1,0,1) (0,1,2). For the purposes of this and previous surveys, the ability of the software to find the optimal model and parameters for the data is characterized as follows: • Automatic forecasting software: Software is labeled as automatic if it recommends both the appropriate model to use on a particular data set and finds the optimal parameters for that model. Automatic software typically searches through multiple potential models to minimize a specific fit metric, such as Akaike Information Criterion (AIC), Normalized Bayesian Information Criterion (BIC) or RMSE; it then recommends a forecast model for the data, gives the model’s optimal parameters, calculates forecasts for a user-specified number of future periods, and gives various summary statistics and graphs. • Semi-automatic forecasting software: The second automation level is called semi-automatic. Such software 52 | ORMS Today
Techniques Available I. Exponential Smoothing
Brown’s Winter’ , Holt’s Two-Pa s Three -Param rameter, eter Me Adaptiv thods e Respo nse Rate Harriso n’s Harm onic Sm Dampe oothing d Trend Program finds Op timal Sm Other fe oothing a statistic tures? (e.g., use [AIC, BIC r c minimiz , root M an specify e S model b d, software cho E, etc.] to be ased on oses ap propriate characte o ristics, e ther [specify] tc.
Import Read Excel 2013, 2010 or older?
2013 an d 2010 2007 an d 2003 If not, w made? hat modificatio (i version, .e. user must sans must be insert a dditiona ve as earlier l inform Addition ation, etc a l .) im port (e.g., sa v, mtw, file formats a V9, etc.) vailable ? Exports Output to Exce Exports l F il e? Session Output? Exports graphic s? (Spe cify form ats) Time-se ries Plo t Scatter Diagram Forecas t Plot Autocorr elation, Partial Autocorr elation, Functio Residua n l Plot Bar, Pie , other B usiness Chart Others (Please specify)
32-bit, 6 4-bit, Platform (e.g.,
(i.e. Win d Mounta ows 8.1, 7, Ma in Unix, etc Lion, Lion, Sn c OS [Maverick ow Leop , .) ard],
Software Product Listing
Software Data Entry Capabilities
etc.) Automa tic Semi-A utomati c Manual Max # o f Observ ations in Time-se If yes, h ries? ow man y?
Operating System(s)
June 2016
asks the user to pick a forecasting model from a menu and some statistic to minimize, and the program then finds the optimal parameters for that model, the forecasts, and various graphs and statistics. • Manual forecasting software: We refer to the third level of automation as manual. Here the user must specify both the model that should be used and the corresponding parameters.The software then finds the forecasts, summary statistics and charts. Note that some products fall into more than one category. For example, if you choose a Box-Jenkins model, the software may find the optimal parameters for that model, but if you specify that Winters’ method be used, the product may require that you manually enter the three smoothing constants. Of the software tools included in the survey, 23 (88 percent) offer semi-automatic forecasting, and 18 (69 percent) offer automatic forecasting capabilities. ORMS Chris Fry (chris@strategicgmgmtsolutions.com) is the founder and managing director of Strategic Management Solutions, an analytics consulting and services firm. Vijay Mehrotra (vmehrotra@usfca.edu) is a professor of business analytics and information systems at the University of San Francisco. The authors thank Gavin Leeper and Craig Volonoski for their contributions to the case study research.
NOTES & REFERENCES 1. The data were collected from NOAA’s database at https://www.ncdc.noaa. gov/cdo-web/search. 2. The El Niño Index tracks temperature changes in the Pacific Ocean. The data we used can be downloaded at https://catalog.data.gov/dataset/climateprediction-center-cpcoceanic-nino-index.
ormstoday.informs.org
Techniques Available
y – –
y y –
y
1
y
y
New Features
Comments
(Since June 2014)
–
– – – – – – – – y Feature selection, k-nearest neighbors, regression trees, etc.
y y Clustering, $995 $30 to y Text mining, ensemble principal to $45, methods, feature components, $1,995 quantity selection, variable logistic discounts screening, Big Data regression, access, partition-onetc. the-fly, many others.
XLMiner, a favorite for teaching, now handles industrial-strength problems and works in Excel, on the Web, or in SDK form.
–
– – – – – – – – y Feature selection, k-nearest neighbors, regression trees, etc.
Text mining, $995, y – clustering, principal
N/A
y New February 2016: Offers developers flexibility/power beyond XLMiner in Excel and XLMiner.com.
XLMiner SDK offers C++, C# and Java developers forecasting, data and text mining power surpassing most R and Python libraries.
– – – – – – – – y Feature selection, k-nearest neighbors, regression trees, etc.
y y
Free
y New January 2016: Cloud SaaS offering, functionality and performance matches desktop XLMiner Pro and Platform.
XLMiner.com offers powerful but easy point-and-click forecasting, data mining and text mining, using only a browser.
seq. replacement, etc.
Unltmd. Depends Forward, – – – on backward, version, stepwise, see www. solver.com/ xlminerlimits
–
seq. replacement, etc.
components, logistic regression, discriminant analysis, classification trees, naive Bayes, etc.
Text mining, clustering, principal components, logistic regression, discriminant analysis, classification trees, naive Bayes, etc.
Free
Can Use r Down load Tria l Versio n?
Unltmd. Depends Forward, – – – on backward, version, stepwise,
Unltmd. Prio ver. Forward, – – – has backward, limits, stepwise,
Free
Educati on
– – – – – – – – – Logistic reg., – y ANOVA, F-test, T-test, z-test, correlation/ covariance, sampling
seq. replacement, etc.
–
Comme rcial
Automa tically fi nd the O On-line ptimal M Help an odel? d Tutori al? Other F eatures (e.g. soft w software are offers fore ca statistic etc.) ally valid sting advice, ates the model,
(e.g. multivariate techniques, etc.)
–
see www. solver.com/ xlminerlimits
y y –
Price
VI. Additional Capabilities
Practical Linear – – – limit reg. 10,000 matches Excel’s Analysis ToolPak
see www. solver.com/ xlminerlimits
y y –
Documentation V. Other Univariate Techniques
Other? (Iplease specify)
IV. Regression Models
Maximu m # of In depend (If yes, h ent Vari ow man ables? y?) Maximu m # of O bservati (If yes, h ons? ow man y?) Other L inear R egressio (If yes, p n Mode lease sp ls? ecify) Nonline ar Regre ssion M Dynamic odels Regress ion Trend A nalysis Specify Trend M odels (i.e. exp onential, S-curve , Weibull, etc.) Spectra l Analys is Kalman Filter Fourier Analysis State S pace M odels Transfe r Functi on Mod Interven el tion Ana lysis Econom etric Mo d els SABL Neural Network s
Moving Average Method Classic s al Deco mpositio Census n Bureau Method s (i.e. X-12 -ARIMA) Program finds Op timal M odel
Time II. Series TimeDecomposition Series III. ARMIA Decomposition (Box-Jenkins)
Platform $1995
$97 per month, Platform $197 per month
y New December 2014: Provides all functions of Excel’s Analysis ToolPak + logistic regression in Excel Online and Google Sheets.
Enables many existing textbooks, tutorials and videos to be used with Excel Online and Google Sheets.
VENDOR DIRECTORY Frontline Systems Inc. P.O. Box 4288 Incline Village, NV 89450 775-831-0300 775-831-0314 info@solver.com www.solver.com
IBM 47 Channel View Road Brighton, East Sussex BN2 6DR UK dunworth@uk.ibm.com
Lumina Decision Systems, Inc. Automatic Forecasting Systems, Inc. P.O. Box 563 Hatboro, PA 19040 215-675-0652 215-675-0652 sales@autobox.com www.autobox.com
Business Forecast Systems, Inc. 68 Leonard Street Belmont, MA 02478 617-484-5050 617-484-9219 info@forecastpro.com www.forecastpro.com
CPDF/Delphus 152 Speedwell Avenue Morristown, NJ 07960 973-267 9269 cpdftraining@delphus.com www.cpdftraining.org
Delphus, Inc. 152 Speedwell Avenue Morristown, NJ 07960 973-267 9269 info@delphus.com www.delphus.com
26010 Highland Way Los Gatos, CA 95033 650-212-1212 info@Lumina.com www.lumina.com
Minitab, Inc. 1829 Pine Hall Rd. State College, PA 16801 814-238-3280 commsales@minitab.com www.minitab.com/
MJC2 33 Wellington Business Park Crowthorne, Berkshire RG45 6LS UK 1344760000 info@mjc2.com http://www.mjc2.com/
NCSS 329 N 1000 E Kaysville, UT 84037 801-546-0445 801-546-3907 sales@ncss.com www.ncss.com
OM Partners
Syncopation Software
Koralenhoeve 23 Wommelgem 2160 Belgium +3236502211 info@ompartners.com www.ompartners.com
6 State Street, Suite 308 Bangor, ME 04401 866-796-2375 866-796-4688 sales@syncopation.com www.syncopation.com
Oracle Corporation
Transpower Corporation
500 Oracle Parkway Redwood Shores, CA 94165 800-392-2999 www.oracle.com
P.O. Box 7132 Penndel, PA 19047 215-355-7011 transpower@aol.com www.transpowercorp.com
RoadMap Technologies, Inc.
Vanguard Software
100 Cummings Center Beverly, MA 01915 978-232-8901 978-232-8916 info@roadmap-tech.com www.roadmap-tech.com
1255 Crescent Green Cary, NC 27518 919-859-4101 sales@vanguardsw.com www.vanguardsw.com/
SAS 100 SAS Campus Drive Cary, NC 27513 919 677-8000 919-677-4444 mike.gilliland@sas.com www.sas.com
Scientific Computing Associates Corp. 212 Lathrop Avenue River Forest, IL 60305 708-771-4567
Statpoint Technologies, Inc. P.O. Box 134 The Plains, VA 20198 540-428-0084 540-428-0089 info@statgraphics.com www.statgraphics.com
June 2016
|
ORMS Today
| 53
Nashville 2016 Come to the “Music City” to learn, share your expertise and experiences, build your professional network, find prospective employers or employees, and reconnect with colleagues and meet new people.
2016 INFORMS ANNUAL MEETING November 13-16 | Nashville, Tennessee
Join 5,000 plus attendees and listen to intriguing plenary and semi-plenary presentations; panel discussions; tutorials; and some of the thousands of oral and poster tracks. INFORMS is looking forward to hosting at the Music City Center & Omni Nashville in the heart of the city. You’ll be steps away from eclectic live music, wonderful restaurants, & Nashville history. We look forward to seeing all of you once again in 2016!
August 1 - Poster Competition Submission Deadline September 1 - Poster Submission Deadline September 16 - All Presenters Must Register October 16 - Early Registration Deadline
http://meetings.informs.org/nashville2016
ormstoday.informs.org
news
Inside News 57 Carnegie Mellon wins UPS Prize 57 MIT team earns IAAA honors 58 Roundtable spring roundup 58 DMDA Workshop 59 Call for nominations 60 People
‘Music City’ tunes up for INFORMS Annual Meeting
Nashville, Tenn., home to more than a hundred music venues such as the legendary Ryman Auditorium (right), will host the 2016 INFORMS Annual Meeting on Nov. 13-16. Images courtesy of Nashville Convention & Visitors Corporation.
By Chanaka Edirisinghe Nashville, Tenn., the Music City, will host the 2016 INFORMS Annual Meeting on Nov. 13-16. While the conference will feature the latest advances in operations research, management sciences and analytics, the host city will offer a sizzling combination of American music, Southern hospitality, unbelievable cuisine and a boundless spectrum of enjoyment. The technical program includes an exciting array of academic and practitioner
invited presentations highlighting several grand challenges facing the world: • Advanced computing as the driver of technological transformation of our society through human-machine interface design and in dealing with climate change, fusion energy, nanotechnology and biotechnology. Topic-related presentations include an opening plenary lecture on cognitive computing by Guruduth Banavar, VP and chief science officer at IBM; a keynote lecture by
61 Winter Simulation Conference 61 Meetings
Jeff Nichols, director (NCCS) at Oak Ridge National Laboratory; and an invited cluster on High Performance Computing organized by Deepak Rajan of the Lawrence Livermore National Laboratory, • Precision agriculture that aims to leverage predictive analytics using real-time data on weather, soil and air quality, crop maturity, etc., to meet the challenges in increasing the global food production in the face of rising population, expected to grow up to 9.2 billion by 2050. Topic-related presentations include an invited cluster organized by Robin Lougee of IBM and Joseph Byrum of Syngenta to bring OR/MS to the forefront of this challenge. • Healthcare issues that have captivated attention, especially during this U.S. election year, with discussions from both policy-making and informatics perspectives. Topic-related presentations include a keynote lecture by Edmund Jackson, chief data scientist and VP of Healthcare Corporation of America; and an invited cluster on O.R.-informed Healthcare Policies, organized by Diwakar Gupta, University of Minnesota. • Exploration of big data and big decisions facing OR/MS researchers and practitioners. Topic-related presentations include a joint plenary lecture by Suvrajeet Sen and Gareth James, University of Southern California; and an invited cluster on Modeling and Methodologies in Big Data, organized by Jiming Peng, University of Houston. • Challenges in identifying and mitigating risk in the financial industry under changing regulations, emerging technologies and heightened corporate responsibility. Topicrelated presentations include an invited cluster on Risk and Compliance organized by Akhtarur Siddique, deputy director (Enterprise Risk and Analysis), Office of the Comptroller of the Currency, U.S. Dept. of the Treasury. Annual Meeting, continued on p. 56
June 2016
|
ORMS Today
| 55
n ews Annual Meeting, continued from p. 55
• The future of global supply chains and the issue of optimizing their efficiencies. Topic-related presentations include a keynote lecture by Jason Murray, VP of World Wide Retail Systems, Amazon; and the IFORS distinguished keynote lecture on hard practical optimization problems that deal with routing bidirectional traffic by Ralph Möhring, Berlin University of Technology.
But Wait, There’s More … Other invited clusters will focus on such themes as additive/advanced manufacturing, physical Internet, energy systems management and entertainment analytics. In addition, the conference will offer a unique opportunity to celebrate the Omega-Rho 40th Year Anniversary through a plenary lecture given by an expert panel consisting of Alfred Blumstein (CMU), John Birge (Chicago), Ralph Keeney (Duke) and John Little (MIT). Throughout the conference, a group of eight academic and six practitioner speakers will give a series of 90-minute tutorial lectures that will be collected as a written volume. The tutorials, a must-make for beginning and advanced researchers alike, will cover the following high-impact research themes: • Optimization frontiers: reviews of stochastic optimization as the science of sequential decision-making under uncertainty, with applications to asset-liability management and Markov decision processes; optimal learning when information is expensive; robust multi-objective optimization theory and applications in engineering, business and management. • Network modeling: systemic risk due to complex dependence structure of interactions among individual components, including banks, financial services providers and regulators – with tutorials in network sampling, resilience under contagion and analysis under behavior of multiple autonomous agents for online social networks and economic and financial markets. • Risk modeling and decisions: review of mathematical finance with emphasis on the need to avoid and rescind destructive deployment of financial risk models; dealing with unstructured data from corporate filings, expert reports and news headlines for financial text mining for risk factors; valuation and hedging of risk in energy portfolios. 56 | ORMS Today
|
June 2016
• Big data: reviews of dimension reduction techniques, as well as data access methods for efficient analytics. • Healthcare & big projects: review of clinical and health sciences research, rooted in empirical evidence, and the role of analytics in addressing health risks of populations; and an expanded review of research and teaching opportunities in project management, a global economic activity valued at $12 trillion annually. Last but not least, the conference will include many clusters and tracks organized by the sponsored societies within INFORMS, as well as many contributed presentations. The INFORMS Roundtable, Analytics Society, CPMS, Railway Applications Section and other practice-related INFORMS societies will collaborate on a special practice track, while presentations of the finalists for the Daniel H. Wagner Prize for Excellence in Operations Research Practice will serve as a Tuesday afternoon keynote. If you have missed the abstract submission deadline for oral presentations, the window is still open to submit poster presentations – either to participate in the poster competition or just to present in one of the poster sessions. These run on Monday and Tuesday from 12:30-2:30 p.m. New this year is the “E-Poster Walk,” an electronic poster displayed digitally. A panel of judges will review the posters entered in the competition and prizes will be awarded the winners.
Something for Everyone The Annual Meeting will take place in the Music City Center, the recently built state-of-the-art convention center, and the adjacent Nashville Omni Hotel. With many downtown hotels within walking distance to the convention center, the attendees will find suitable accommodations to fit all budgets. We look forward to another record-setting number of more than 5,000 attendees at INFORMS 2016. I am grateful to the organizing committee members who have steadfastly volunteered to put together a well-orchestrated and comprehensive program for 2016. They are Shabbir Ahmed (invited), James Primbs (plenaries), Aparna Gupta and Agostino Capponi (tutorials), Melissa Bowers and Oleg Shylo (posters), Mingzhou Jin and
Sean Willems (practice), James Ostrowski (sponsored), Scott Mason and Justin Yates (contributed), Christine Vossler and Anahita Khojandi (arrangements), and the program chair, Bogdan Bichescu. My special appreciation goes to the highly professional and efficient INFORMS staff members, who are always there to help and guide the organizing committee. I would be remiss not to mention that Nashville offers everything from an electrifying multi-genre music scene, awardwinning cuisine, historic homes, world-class art, a myriad of attractions, unique shopping, college and professional sports and more. There are more than 120 live music venues across the city; you’ll catch pickers and songwriters all over town, in places such as the bluegrass venue Station Inn, the rock venue Exit/In, the honky-tonks on Broadway, the song-centered Bluebird Cafe or the legendary Ryman Auditorium. Nashville is also home to many attractions, from the Grand Ole Opry to the world-renowned Country Music Hall of Fame and Museum. Opened in May 2013, the Johnny Cash Museum and the Musicians Hall of Fame and Museum add to the lineup. Nashville is also rich in visual and fine arts. The Frist Center for the Visual Arts is housed in an exquisitely converted Art Deco post office, the Cheekwood Botanical Garden and Museum of Art houses an extensive art gallery, and art crawls, studios and art galleries are all around town. Dubbed “The Coolest, Tastiest City in the South” by Bon Appétit magazine. Nashville’s creative spirit has infiltrated into its kitchens, turning them into chef’s studios. From Southern fare to haute cuisine to quite literally everything in between, Nashville’s palate offers it all. Simply put, this is an annual meeting you cannot afford to miss. With the usually perfect mild weather expected during the meeting dates, join us at INFORMS-Nashville to share expertise and experiences, work and network toward a brighter future. For more information, see: http:// meetings2.informs.org/wordpress/ nashville2016/ ORMS Chanaka Edirisinghe, professor, Lally School of Management, Rensselaer Polytechnic Institute, is the general chair of the 2016 INFORMS Annual Meeting in Nashville, Tenn.
ormstoday.informs.org
ormstoday.informs.org
Carnegie Mellon schools receives UPS George D. Smith Prize The School of Information Systems & Management and the School of Public Policy and Management at the H. John Heinz III College, Carnegie Mellon University, received the 2016 UPS George D. Smith Prize from INFORMS. The award recognizes excellence in preparing students to become practitioners of analytics and operations research. The award was presented at the 2016 INFORMS Conference on Business Analytics & Operations Research held April 10-12 in Orlando, Fla. The other finalists included the Institute for Advanced Analytics at North Carolina State University and the Operations Research Program at the U.S. Air Force Academy. The UPS George D. Smith Prize is awarded to an academic department or program for effective and innovative preparation of students to be good practitioners of operations research, management science or analytics. Established in 2011, the award is named in memory of the late UPS chief executive officer who was a patron of operations researchers at the Fortune 500 company. “We are proud of the work this year’s INFORMS George D. Smith Prize recipient, Heinz College at Carnegie Mellon University, is doing to develop the next generation of operations research and analytics practitioners,” said Chuck Holland, UPS vice president of engineering. “At a time when world leaders are struggling to find answers to complex problems – global trade, emerging markets, poverty, and hunger among many others, Operations research is a discipline they should turn to for solutions. These O.R. and analytics students are the key to a better future.” Heinz College’s unique and effective analytical education, experiential learning activities and successful collaboration between students and partner organizations for capstone projects played an important role in its winning this award. Those collaborations allow students to put skills into practice. Through experiential learning opportunities like capstone projects, internships and apprenticeships, Heinz students help to research and develop solutions to
some of society’s most pressing challenges. “We are honored and excited to receive this prestigious award,” said Heinz College Dean Ramayya Krishnan. “As the School founded by William W. Cooper, a legendary operations researcher, analytic thinking, appropriate use of technology and a deep interest CMU faculty members Al Blumstein (front, center, holding trophy), in societal problem-solving are Ramayya Krishnan (second from right) and others celebrate the embedded in our DNA.” UPS Prize. “INFORMS has a long and rich tradition of honoring the very best of those awards. We congratulate Carnegie in operations research and analytics through Mellon University’s Heinz College for winan array of awards, conferences and publica- ning the 2016 Smith Prize.” tions,” said Melissa Moore, INFORMS execThe UPS George D. Smith Prize inutive director. “The Smith Prize is a key part cludes a $10,000 cash award. ORMS
MIT-led team wins IAAA competition A team from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), in collaboration with the Lahey Hospital and Medical Center in Burlington, Mass., won the 2016 Innovative Applications in Analytics Award (IAAA) for research about how digitally connected tools could be used to help diagnose brain disorders. The flagship competition of the Analytics Society of INFORMS, the award recognizes creative and unique applications of a combination of analytical techniques in a new area. The winning entry, “An Analytics Approach to the Clock Drawing Test for Cognitive Impairment,” was selected by a panel of judges following a series of presentations at the INFORMS Conference on Business Analytics & Operations Research in Orlando, Fla. Last year, professors Randall Davis and Cynthia Rudin of MIT, then-grad student William Souillard-Mandar and Dr. Dana Penney of the Lahey Hospital and Medical Center in Burlington, Mass., published a paper demonstrating a set of analytical and machine-learning techniques that, when coupled with existing hardware, open up
the possibility of detecting disorders such as dementia earlier than ever before. For several decades, doctors have screened for conditions including Parkinson’s and Alzheimer’s with the clock drawing test, which asks subjects to draw an analog clock-face showing a specified time, and to copy a pre-drawn clock. Dr. Penney has been the clinical lead on the project, with test collection going on at Lahey and partnering medical institutions. The project uses a commercially available digitizing ballpoint pen that measures its position on the paper 75 times a second via a camera built into the pen. This provides data that are far more precise than can be measured on an ordinary drawing, and captures timing information. Davis and Penney created software that computes a large number of novel measurements on this data, enabling the analysis of both the end product -- the drawing – and the process that produced it, i.e., all of the subject’s movements and hesitations. In addition to a cash award of $2,000, the team will be recognized at the INFORMS Annual Meeting in Nashville, Tenn., where it will give a reprise presentation. ORMS June 2016
|
ORMS Today
| 57
n ews Spring Roundtable Meeting in Orlando By Dave Hunter The spring 2016 Roundtable Meeting was held on April 9-10 at the Hyatt Regency Grand Cypress in Orlando, Fla. The meeting theme was “Roundtable Companies” and featured four talks by Roundtable member organizations describing their groups and projects. The meeting also featured a two-hour discussion session titled “What does the Roundtable think,” which included breakout sessions to discuss several questions relevant to our industries. An outing was held on Saturday afternoon to Escapology in Orlando, which provided an excellent networking opportunity. The participating Roundtable members and guests split into four teams and each team entered a different themed room. Two of the four groups were able to successfully solve all the puzzles and escape in time! Much cheering could be heard from the triumphant Roundtable members as they exited the room. The other two groups were not as fortunate and did not make it out before time expired. Following Roundtable introductions, the main program began on Sunday morning with a presentation by Jennifer Rausch. Rausch leads the Operations Research and Analysis team at Jeppesen, which is a Boeing subsidiary. She showed an interesting video on the history of Jeppesen and then provided examples of some current analytics being developed by her team. The next speakers were Larry Stone (chief scientist) and Van Gurley (senior vice president) from Metron. Stone spoke first and discussed the development of probability maps and showed how they were used to locate the S.S. Central America, which sank in 1857 taking millions of dollars of gold to the bottom of the ocean. Gurley followed by discussing several of Metron’s current projects. He also provided an overview of his 26-year Navy career – including pictures! Following lunch, the third speaker was Tim Merkle, who leads the Global Advanced Analytics team at Steelcase. Merkle is also a veteran, having served nine years in the U.S. Marine Corps before joining Steelcase. He shared his perspectives on the analytics journey to develop and execute successful 58 | ORMS Today
|
June 2016
advanced analytics initiatives across the Steelcase enterprise. The final speaker of the meeting was Margery Connor from Chevron. Connor has enjoyed a 28+ career at Chevron, and she currently leads the new Modeling and Analytics Center of Excellence in IT. Connor noted that while Chevron has a long history of applying operations research and advanced analytics in pockets of the company, she is now very focused on the enterprise value of applying advanced analytics. She shared the journey to date, some lessons learned and
the next steps during these turbulent times for the energy industry. Following the member presentations, Roundtable President William Browning conducted the annual Roundtable business meeting. He provided a status update on the Roundtable membership and finances and answered questions from the Roundtable. Jeff Winters, Roundtable VP meetings, followed and spoke about planning for future Roundtable meetings. The spring meeting concluded with a farewell networking dinner at Hemmingway’s restaurant. ORMS
INFORMS Workshop on Data Mining & Decision Analytics The Data Mining, Analytics, Artificial Intelligence, Quality, Statistics & Reliability and Multi-Criteria Decision Making sections of INFORMS, along with the Health Applications Society of INFORMS, are organizing the 2016 INFORMS Workshop on Data Mining and Decision Analytics to be held Nov. 12 in Nashville, Tenn., in conjunction with the 2016 INFORMS Annual Conference. Paper submissions are welcomed. Submissions, no more than 15 pages in length, should be sent to ellen.tralongo@ informs.org. The deadline is June 30. The DMDA Workshop will include a Best Paper Presentation Contest. Authors of all papers accepted after a rigorous two-round review process and after an in-person presentation at the workshop are eligible for the contest. The contest, based on both the content of the paper as well as the presentation, will be judged by a panel of workshop co-chairs and session chairs. Any topic related to data mining & decision analytics will be considered. In addition to a name-engraved plaque, the top three finishers will receive monetary awards. Workshop participants are also encouraged to submit extended and enriched manuscripts for a special issue of the journal Decision Sciences on the topic of Data Mining & Decision Analytics. For more information, contact the special issue guest editor Asil Oztekin (Asil_Oztekin@uml.edu).
The workshop will feature an editors’ panel on the topic, “How to publish in toptier journals,” in which representatives of leading journals in the field of data mining and decision analytics will make short presentations. A Q&A session will follow, in which those in attendance will have the opportunity to ask the editors questions. Keynote speakers include: Amit Basu, the Carr P. Collins Chair of Management Information Sciences at the Edwin L. Cox School of Business at Southern Methodist University; David L. Olson, the James & H.K. Stuart Professor in MIS and Chancellor’s Professor at the University of Nebraska; Nick Street, professor, departmental executive officer and Henry B. Tippie Research Professor in the Management Sciences Department at the University of Iowa; and Olivia Sheng, Presidential Professor and Emma Eccles Jones Presidential Chair of Information Systems at the David Eccles School of Business. ORMS For more information, contact the workshop co-chairs: Asil Oztekin (Asil_Oztekin@uml. edu), University of Massachusetts Lowell; Cem Iyigun (Iyigun@metu.edu.tr), Middle East Technical University; and Ramin Moghaddess (Ramin@miami.edu), University of Miami; and visit http://meetings2.informs.org/wordpress/ nashville2016/data-mining-decision-analytics/
ormstoday.informs.org
ormstoday.informs.org
Call for Nominations: INFORMS Fellows INFORMS Fellows are examples of outstanding lifetime achievement in operations research and the management sciences. They have demonstrated exceptional accomplishments and made significant contributions to the advancement of OR/MS over a period of time. A nominee for the Fellow Award must be a current full member or retired member of INFORMS and must have been a full or retired member for at least 10 years, not necessarily consecutive. However, longterm membership in INFORMS or ORSA or TIMS more than 10 years ago may mitigate a membership gap of a year or two during the last 10 at the discretion of the Organizing Committee. Officers and other Board members of INFORMS, and INFORMS staff members, may not nominate or be nominees while holding such positions, unless the individual has been ineligible to nominate or be a nominee for five or more consecutive years as a result of this restriction. Members of the Fellow Selection Committee may not nominate or serve as references. The contributions of a nominee will be evaluated in each of the following five categories; contributions must be outstanding in at least one category: • Research, includes theoretical, empirical or computational innovations in OR/MS; • Practice, includes substantial application of OR/MS to significant practical problems; • Management, includes significant responsibility for and direction of the development and application of OR/MS techniques and knowledge, within an organization of any type (e.g., academic, for-profit, nonprofit, governmental, military, healthcare), over an extended period of time, that have had a major impact internal and/ or external to the organization; • Education, includes activities that had significant impact on the growth and development of OR/MS education; • Service, includes significant work over an extended period of time on
behalf of INFORMS and its functions and/or significant contributions that advanced the stature and recognition of the OR/MS profession. A nomination shall be prepared by a full or a retired member of INFORMS (the nominator). The nominator is responsible for securing exactly three persons to serve as references for the nominee. At least one of the references must be an INFORMS Fellow and the remaining references should be current regular or retired members of INFORMS. Members of the Selection Committee may not serve as references.
The nominator and at most one of the reference letters can be from the same institution as the nominee. A maximum of four reference letters, including the letter from the nominator, may be submitted. For more details on eligibility, criteria, and nomination and selection procedures, visit https://www.informs.org/ Connect-with-People/Fellows. Complete nominations are due by July 15, 2016, to: INFORMS, Attn: Fellows Nominations, 5521 Research Park Drive, Suite 200, Catonsville, MD 21228 USA; email: Tracy.Cahall@informs.org. ORMS
Call for Nominations:
Editor-in-chief, Organization Science The term of Professor Zur Shapira as editor-in-chief of the journal Organization Science will expire on Dec. 31. Zur has decided not to seek a second term so we need to begin the process of finding his successor. We will celebrate Zur’s exceptional contributions to the journal at a later date. Based on recommendations f r o m t h e I N F O R M S Pu b l i c a t i o n s Committee, the president of INFORMS has appointed a committee to conduct a full search for a new editor-in-chief. The committee intends to propose a candidate for approval by INFORMS no later than Sept. 30. All members of INFORMS are invited to participate in this process. The committee seeks your opinions and comments on (1) the current state of the journal; (2) recommendations for change, if any; and (3) candidates for editor-in-chief. The deadline for nominations is June 30. Qualifications for the editor-in-chief of Organization Science include: • a demonstrated interest in a broad range of topics in the field; • a demonstrated record of excellence in research in the field; • dedication and enthusiasm for Organization Science; • significant editorial experience; • vision of the role of scholarly
publications in the electronic and open-access age; • commitment to the workload involved; • ability to effectively and efficiently manage the editorial process; and • strong communication skills and demonstrated ability to work with people. Further information about the journal is available at the Organization Science website: http://pubsonline.informs.org/journal/ orsc. Members of the search committee are Felipe Caro (UCLA, and committee liaison from the INFORMS Publications Committee), Martine Haas (University of Pennsylvania), Kyle Mayer (University of Southern California), Phanish Puranam (INSEAD) and Linda Argote (Carnegie Mellon University, and committee chair). Submit comments, recommendations and nominations (including self-nominations) by June 30. The information may be sent (preferably by email) to the chair of the Organization Science Editorial Search Committee: Prof. Linda Argote, Tepper School of Business, Carnegie Mellon University, Pittsburgh, PA 15213; argote@ cmu.edu. ORMS June 2016
|
ORMS Today
| 59
n ews People Jonathan Caulkins, the H. Guyford Stever Professor of Operations Research and Public Policy at the H. John Heinz III College at Carnegie Mellon UniJonathan Caulkins versity, was named a University Professor at CMU, the highest designation a faculty member can achieve at the university. “University Professors are distinguished by international recognition and for their contributions to education, artistic creativity and/or research,” said CMU Provost Farnam Jahanian. An INFORMS Fellow and a 2010 recipient of the INFORMS President’s Award, Caulkins is internationally known for his research in public policy issues, particularly modeling the effectiveness of interventions related to drugs, crime, violence, delinquency and prevention. He is a former co-director of RAND’s Drug Policy Research Center. At the Heinz School, he has served as director of the Masters of Science in Public Policy and Management (MSPPM) program and as interim associate dean for faculty. He taught at CMU’s Qatar campus from 2005 to 2011. A member of the National Academy of Engineering, Caulkins has authored more than 130 articles and 10 books. He earned a Ph.D. in operations research from MIT in 1990, the same year he joined the faculty at CMU. Janis Terpenny, professor and department head with the Harold and Inge Marcus Department of Industrial and Manufacturing Engineering, was inducted into the Virginia Tech Industrial and Systems Engineering (ISE) Academy of Distinguished Alumni. The Academy was established in 1993 and includes a “group of recognized leaders in the field who have served their profession and society with distinction.” 60 | ORMS Today
|
June 2016
A member of INFORMS, Terpenny earned her master’s and doctoral degrees in ISE from Virginia Tech in 1981 and 1996, respectively. “I am humbled Janis Terpenny and honored to have been selected to join such an inspiring group of individuals, dedicated to leading the profession into the future and serving society,” Terpenny said. “I will be forever grateful to my master’s and doctoral advisors at Virginia Tech, Dr. Robert P. Davis and Dr. Paul E. Torgersen … my heroes, who were outstanding educators and thought leaders, wonderful mentors and role models for me and for so many others.” Terpenny is also the director of the Center for e-Design, a National Science Foundation industry/university cooperative research center that brings eight universities and more than 30 industry members together in solving pressing problems associated with the design, manufacture, delivery and sustainment of innovative, high-quality, lower cost products. Terpenny was named a Fellow of the Institute of Industrial Engineers in 2010 and of the American Society of Mechanical Engineers in 2012. Her research and teaching interests include engineering design process and methods and engineering design education, including engineering economics, intelligent and integrated design systems and systems analysis/systems engineering. A University of Alabama student team from the Culverhouse College of Commerce recently won the 2016 Dow Big Data Challenge. Team members Bryant Cassidey of Mobile, Ala., and Ford McDermott of Atlanta are master’s degree students in operations management, specializing in decision analytics at Culverhouse.
Ford McDermott (left) and Bryant Cassidey
They competed with some 20 teams composed of up to two graduate students and one undergraduate student. Dow Chemical formulated a problem from its daily operations and provided the data to the teams in the competition. The problem represented a Dow business segment either in industrial engineering, operations research, supply chain or management sciences. The students analyzed the data, developed solutions and presented their solutions to Dow executives for judging. “We are proud of Bryant and Ford,” said Dr. Burcu Keskin, UA associate professor of operations management and an active member of INFORMS. “This is the second year that a Culverhouse team has placed in this competition. The success of our students is evidence of the quality of our analytics program at Culverhouse.” The Culverhouse team placed second in last year’s Dow Challenge. Harriet Nembhard led a team of faculty, staff and students from the Penn State Center for Integrated Healthcare Delivery Systems (CIHDS) that participated at the ormstoday.informs.org
ormstoday.informs.org
Meetings National Science Foundation Center for Health Organization Transformation (CHOT) Industry Advisory Board spring meeting in April at the Texas Children’s Hospital Harriet Nembhard in Houston. Nembhard is a professor of industrial engineering at Penn State and a member of INFORMS. The five proposals the CHIDS team presented in Houston included: 1. “Value-based Care: Challenges in a Changing Program” 2. “Integration of Genomic Data for Precision Health Decision Support” 3. “Sensing Systems for Personalized Telehealth Wellness Management” 4. “A Person-Centered Approach for Individuals with Multiple Chronic Conditions” 5. “Demand Management for Community Paramedicine” A s a N S F I n d u s t r y- U n i v e r s i t y Cooperative Research Center, CHOT f o l l o w s a m o d e l o f a n i n d u s t r yacademic partnership that has benefited industry-focused research across more than 50 disciplines. CHOT creates a safe, mutually beneficial, cooperative environment in which leading healthcare industry members can come together in collaboration to support important transformation initiatives addressing health organization management and services; examine the implementation of transformational strategies; partner with healthcare management researchers to improve such initiatives and strategies; participate in research in a cost-effective manner; and play a critical role in shaping the education of future healthcare leaders as managers, engineers and health professionals. ORMS
Go to www.informs.org/Conf for a searchable INFORMS Conference Calendar.
INFORMS Annual & International Meetings
INFORMS Community Meetings
June 12-15, 2016
June 16-18, 2016
2016 INFORMS International Meeting
2016 INFORMS Marketing Science Conference
Hilton Waikoloa Village Waikoloa, Hawaii Chair: Saif Benjaafar, University of Minnesota http://meetings.informs.org/2016international
Shanghai, China Chair: Icey Han, Fudan University http://www.fdsm.fudan.edu.cn/marketingscience2016/
Nov. 13-16, 2016
June 30-July 1, 2016
INFORMS Annual Meeting
MSOM Conference
Music City Center & Omni Nashville Nashville, Tenn. Chair: Chanaka Edirisinghe, RPI http://meetings.informs.org/nashville2016
April 2-4, 2017
University of Auckland Business School Auckland, New Zealand Co-chairs: Tava Olsen and David Robb, University of Auckland Business School http://www.cscm.auckland.ac.nz/2016-msom-conference
INFORMS Conference on Business Analytics & Operations Research
Dec. 11-14, 2016
Caesars Palace, Las Vegas Las Vegas, Nevada
Crystal Gateway Marriott Arlington, Va. Chair: Todd Huschka, Mayo Clinic
Winter Simulation Conference
Oct. 22-25, 2017
INFORMS Annual Meeting George R. Brown Convention Center & Hilton Americas Houston, Texas Chair: William Klimack, Chevron
2016 Winter Simulation Conference The 2016 Winter Simulation Conference (WSC 2016) will be held Dec. 11-14 in Arlington, Va., at the Crystal Gateway Marriott. The hotel is located minutes from downtown Washington, D.C., and less than a mile from Reagan National Airport. The METRO is connected to the hotel, which makes exploring the area’s monuments, museums, shopping and restaurants very convenient. The WSC is the premier international forum for disseminating recent advances in the field of dynamic systems modeling and simulation. In addition to a technical program of unsurpassed scope and quality, WSC provides the central meeting place for simulation practitioners, researchers and vendors. Research in modeling and simulation is
propelled by fostering cross-fertilization between various disciplines. The theme for WSC 2016 is “Simulating Complex Service Systems.” The theme emphasizes the increasingly complex engineered and human systems in highly connected environments, the availability of data to help model such systems, technological advances which continue to push the limits of computation, and conceptual and mathematical advances that help make sense of complex systems. These forces enable more informed decisions. WSC 2016 is sponsored by ACM/ SIGSIM, IIE, INFORMS-SIM and SCS, along with technical co-sponsors ASA, ASIM, EEE/SMC and NIST. For more information or to register for the conference, visit wintersim.org. ORMS
Check out the latest issue of Analytics magazine: http://www.analytics-magazine.org/
June 2016
|
ORMS Today
| 61
Industry News LINDO announces new releases of optimization modeling tools LINDO Systems is shipping new releases of its family of optimization modeling tools What’sBest! 14, LINGO 16 and LINDO API 10. The new versions include a wide range of performance enhancements and new features. Faster solutions on linear models with improved Simplex solver: Enhancements to the Simplex solvers boost performance on large linear models. Large models solve an average of 35 percent faster using primal simplex and 20 percent faster for dual simplex. Improved integer solver with new features: A new optimization mode has been introduced to ensure reproducibility of runs. Users can investigate alternative optima more quickly. Enhancements to the K-Best algorithm allow finding K-Best solutions in little more time than finding one solution. Models with knapsack constraints and block structures solve faster using new heuristic algorithms. New preprocessing level tightens variable bounds for better performance on classes of nonlinear models. Enhanced stochastic solver: Large linear multistage SP instances solve 60 percent faster with improved cut management for Nested Benders Decomposition Method. Enhancements improve handling of multistage SP models that do not have fullrecourse. Extensions to the parser allow the use of arbitrarily complex functions of stochastic parameters. Improved global solver: Performance of the global solver has been dramatically improved on classes of quadratic problems, particularly non-convex quadratic problems rejected by other solvers, or otherwise solvable only slowly to a local optimum by traditional NLP solvers. The new releases solve some previously intractable problems to global optimality, especially financial portfolio models with minimum buy quantities, and/or limit on number of instruments at nonzero level. Along with dramatically faster, more robust performance on many linearized models using functions like @MAX( ), @ MIN( ), @ABS( ), x*z where z = 0 or 1, etc., the global solver incorporates a new bound tightening process to the linearization procedure and improves solvability of linearized models. New scenario viewer in What’sBest!: Powerful new feature to view different 62 | ORMS Today
|
June 2016
solutions on models with integer variables and stochastic programming models. On integer models you can use the scenario viewer to browse through alternative optima and near optimal solutions. On stochastic programming models you can scroll through different scenarios of the scenario tree. LINGO offers native Macintosh and Linux support: The new Mac and Linux versions of LINGO now feature the same graphical user interface as the popular Windows version. AIMMS, customers featured in Inbound Logistics magazine AIMMS and AIMMS customers JBS and Nampak were featured in Inbound Logistics magazine. The article explores the opportunities and benefits of supply chain modeling, drawing on insights from AIMMS’ Marcel Mourits and Celso Batista (coordinator of planning and control at JBS in Brazil) among others. As illustrated in the article, companies are starting to embrace more flexible software options that can adapt to their specific needs instead of off-the-shelf solutions. With supply chain modeling, they can explore different business scenarios, anticipate future disruptions and ultimately make better decisions. Excerpts from the article: “Companies can start playing with reality without actually affecting reality,” explains Marcel Mourits, supply chain optimization lead at AIMMS, a global company offering integrated multidimensional modeling software. “They can start asking the important question: What if? What if we close this warehouse? What if revenue grows by 20 percent? What if we start using multimodal shipping instead of relying only on trucking?” One key quality of the modeling software is the way it forces businesses to challenge the kind of assumptions and conventional wisdom that can lead to lazy thinking. No matter how JBS uses modeling to peer into the future, the process and tools will challenge the company and force it to scrutinize itself in a way spreadsheets never did, Batista says. “Modeling software makes companies question their thinking and get creative,” he says. “It’s not the kind of system that simply provides production plans, and all the user has to do is push ‘print.’”
FICO, Guidepost Solutions team up to fight anti-financial crime Analytic software firm FICO and risk consultancy Guidepost Solutions LLC today announced that they have entered into a partnership to help provide companies worldwide with a complete suite of anti-financial crime solutions. In this new agreement, Guidepost Solutions will offer the FICO TONBELLER Siron Anti-Financial Crime Solutions for anti-money laundering, regulatory compliance and tax compliance with services developed for its clients around the world, beginning with efforts in the U.S. and France. “The FICO TONBELLER integrated suite of products fills a gap in the market for global companies with diverse business models and changing regulatory needs,” stated Guidepost Solutions Chief Executive Officer Julie Myers Wood.“Working in partnership, Guidepost Solutions and FICO are uniquely positioned to build effective compliance programs that provide clients with a 360-degree view of their customers and enhance their ability to meet regulatory compliance requirements.” “The consultants at Guidepost Solutions are respected throughout the risk industry,” said Torsten Mayer, vice president for compliance solutions at FICO. “They understand compliance from multiple perspectives, and many have served as independent compliance consultants or corporate integrity monitors of Fortune 500 companies. Our technology will help their clients manage a rapidly changing regulatory landscape at a lower cost and with fewer staff.” Guidepost Solutions routinely assists clients in designing, developing, implementing and testing compliance programs. Its professionals have been providing independent compliance consulting and monitoring services for more than three decades for public and private companies, including Fortune 500 companies and global entities; domestic and international banks; law firms; governmental and quasi-governmental agencies; and not-for-profit organizations. The FICO TONBELLER Siron AntiFinancial Crime Solutions suite consists of flexible and highly integrated software modules for anti-money laundering, tax compliance, counter-terrorism financing and other issues. ORMS ormstoday.informs.org
SPECIAL ADVERTISING SECTION | View Classifieds Online at: http://www.orms-today.org
CLASSIFIEDS
advertiser index Page #
ADVERTISER
C2
(Paragon Decision Technology)
13
Dynamic Ideas, LLC
E-mail & Web Page
AIMMS info@aimms.com www.aimms.com
orders@altlog.net www.dynamic-ideas.com
11 FICO
Jerry S. Dobrovolny Chair in System Engineering and Design
FICOAmericas@fico.com www.fico.com/optimization
3, Frontline Systems, Inc. 5, info@solver.com 7 www.solver.com BC
GAMS Development Corp.
9
Gurobi Optimization
($2M endowed chair)
sales@gams.com www.gams.com
The Department of Industrial and Enterprise Systems Engineering at the University of Illinois at Urbana-Champaign invites applications for a full-time faculty position at the rank of full professor in the area of large scale systems engineering and design. We are seeking a preeminent scholar who uses a multidisciplinary approach to the study of systems engineering and design. We are especially interested in a visionary leader whose research agenda will contribute to our campus excellence.
713.87.9341 info@gurobi.com www.gurobi.com
14, 15, 17, INFORMS 33, 39, 47, informs@informs.org 54 meetings@informs.org www.informs.org
1
LINDO Systems, Inc.
C3
Optimization Direct, Inc.
info@lindo.com www.lindo.com
alkis@industrialgorithms.com www.optimizationdirect.com
H T T P : / / W W W. A N A LY T I C S - M A G A Z I N E . O R G
DRIVING BETTER BUSINESS DECISIONS
MAY/ J U N E 2016 BROUGHT TO YOU BY:
MANKIND & MACHINES
Applicants at the rank of full professor will be considered. Successful candidates are expected to direct graduate students in research, teach in the undergraduate and graduate programs, and develop a strong externally-funded research program in the area of systems engineering and design. Salary will be commensurate with qualifications and experience. Candidates must have a PhD in Industrial Engineering, Systems Engineering, Mechanical Engineering, or a closely related discipline. For complete position announcement and application instructions, see http://jobs.illinois.edu. Review of applications will be ongoing, and will continue until the position is filled. The proposed start date is August 16, 2016. Questions should be referred to Shawna Graddy, sgraddy@illinois.edu, (217) 244-8788.
• It takes both to get most from big data & analytics • Will computers replace knowledge workers?
ALSO INSIDE: • Self-service analytics: requisite skills • Making wise decisions with the IoT • Cloud-supported machine learning • Corporate profile: Steelcase Inc.
Executive Edge Mu Sigma strategy guru Tom Pohlmann on analytics job one: closing the data-todecisions gap
Check out the May/June 2016 Issue of ANALYTICS Now Available at: www.analytics-magazine.org
The University of Illinois conducts criminal background checks on all job candidates upon acceptance of a contingent offer. Illinois is a EEO Employer/Vet/Disabled (www.inclusiveillinois.illinois.edu) and committed to a family-friendly environment (http://provost.illinois.edu/worklife/index.html).
June 2016
|
ORMS Today
| 63
ORacle
Doug Samuelson
samuelsondoug@yahoo.com
The Orator’s Parable Ben and Brett, two OR/MS analysts, had sat quietly through the presentation by a senior government official and potential research sponsor at a professional conference. “What did you think of that?” Brett asked Ben. “He seemed to have a lot to say.” “Yes he did,” Ben concurred, “but that’s the problem. He had too much to say and tried to cram all of it onto those slides. Made it awfully hard to follow and digest, didn’t it?” Brett nodded. “I wonder,” Ben continued, “whether that guy would let me sell him a few speech coaching sessions. Heaven knows he needs them.” “You coach speakers?” Brett asked, surprised. “Yeah, believe it or not,” Ben affirmed. “I took state in oratory back in high school. I’ve done some speechwriting and speech coaching back when I was in government agencies. And, of course, plenty of briefings and conference presentations and classroom lectures myself. So, yes, I could help him. It wouldn’t be that hard. He literally doesn’t know the first thing about pubic speaking.” “And what is that?” Brett inquired, even more surprised. “Forget about what you want to say,” Ben smiled. “What?” Brett blurted. “Focusing on what you want to say is a distraction, or worse,” Ben explained. “You want to focus on what you want the audience to remember.” Brett took a moment to absorb this. “This has several advantages,” Ben elaborated. “First of all, you figure out who your audience is and try to meet them where they are – that is, start with what you’re pretty sure they know and believe and build from there.You emphasize major points more, repeat them several times, and make sure they’re clearly 64 | ORMS Today
|
June 2016
and simply – memorably – worded.Write your conclusion first, then your introduction, and make sure the conclusion summarizes your key point in a short, catchy phrase that they’ll remember if they remember nothing else. Then you build from introduction to conclusion in the order that’s easiest to follow, not the order that makes the most sense to you. This also means you time it so that if a session chair might cut you off, you make sure to get the conclusion in even if your delivery runs longer than you intended.” Ben went on, “Use examples close to their experience, if you can. Use examples to back up your most important points and try to keep the descriptions limited to what does back up those points – extra information is distracting and confusing, not helpful.You can provide backup slides or additional reading material to add detail, but the main spoken and slide presentation should be terse, no more than five bullet points per slide, in a large font. And I usually go with yellow or orange print on a blue background, because that’s the easiest color scheme for people to see. Black on white is much harder to read.” Brett listened attentively. “Humor is good,” Ben said, “but you have to be careful. A speakers’ bureau rep pointed out to me that if you have a thousand people in the room, and you tell a blue joke that offends three of them and amuses the rest, the speakers’ bureau will get one laudatory letter from the event organizer and three complaint letters. So your mail is running three-to-one against you. “There’s a similar insight about contests,” Ben continued. “There you have to change your focus from what you want the audience to remember and like to what you want the judges to like. If there are a few hundred people in the room and most of them like your speech, but three of the five judges don’t, you lose! And one
quick way to do just that is to challenge your audience. Instead, you tell the judges exactly what they already passionately believe, in a slightly novel way.You tell the American Legion that communism is bad; you tell the Chamber of Commerce that business is praiseworthy. “And if the judges are Toastmasters,” Ben went on, “you get a Toastmasters ballot and make sure to use the rhetorical devices they include in their scoring system. Toastmasters is great for helping scared beginners make competent speeches, just like Arthur Murray is good at getting beginners to dance. But when you want to get good, you have to change your style so that it’s not obvious where and how you learned. So you have to alter your style to fit what the judges are looking for. I gave a lot of thought-provoking speeches my friends and I thought were good – but I didn’t win. After I got coached about how to win, I gave speeches that didn’t impress my friends so much, but I went on quite a winning streak!” “OK. Next week I’m scheduled to give a presentation to a research group that might hire me,” Brett told Ben.“How do I impress them?” Ben laughed. “My dissertation advisor told me how to do that. Research types think what you say is trivial if it’s too easily understood. So identify the key decision-maker – probably the department chair. Let’s say you have an hour, that’s pretty typical. In the first 15 minutes, describe the problem you tackled and why anyone should care. Everyone in the room should understand that. Then in the next 15 minutes, lose the graduate students. In the third 15 minutes, lose everyone except the department chair. In the last 15 minutes, lose him.” “Cynical, but most likely accurate,” Brett chuckled. “It is. And that,” Ben concluded, “helps to explain a lot about academic presentations and publications, doesn’t it?” ORMS Doug Samuelson (samuelsondoug@yahoo. com) is president and chief scientist of InfoLogix, Inc., in Annandale, Va.
ormstoday.informs.org
CPLEX Optimization Studio®. Still the best optimizer and modeler for the energy industry. Now you can get it direct
CPLEX Optimization Studio is well established as the leading, complete optimization software. For years it has proven effective in the energy industry for developing and deploying models and optimizing business decisions. Now there’s a new way to get CPLEX – direct from the optimization industry experts. Find out more at optimizationdirect.com The IBM logo and the IBM Member Business Partner mark are trademarks of International Business Machines Corporation, registered in many jurisdictions worldwide. *IBM ILOG CPLEX Optimization Studio is trademark of International Business Machines Corporation and used with permission.
OPTIMIZATION GENERAL ALGEBRAIC MODELING SYSTEM High-Level Modeling The General Algebraic Modeling System (GAMS) is a high-level modeling system for mathematical programming problems. GAMS is tailored for complex, large-scale modeling applications, and allows you to build large maintainable models that can be adapted quickly to new situations. Models are fully portable from one computer platform to another.
State-of-the-Art Solvers GAMS incorporates all major commercial and academic state-of-the-art solution technologies for a broad range of problem types.
Optimizing Carbon Capture Technologies: The CCSI Optimization Toolset Carbon Capture technologies can significantly reduce atmospheric emissions of CO2 from fossil fuel power plants. A widespread deployment of these technologies is necessary to significantly reduce greenhouse gas emissions and contribute to a clean energy portfolio. But the deployment is both expensive and time-consuming: bringing such technologies online can take industries between 20 and 30 years. Speeding up this process is the express goal of the Carbon Capture Simulation Initiative (CCSI). Founded by the U.S. Department of Energy in 2011, CCSI is a partnership among national laboratories, industry and academic institutions. High-Level Modeling for the Success of Future Technologies CCSI provides an optimization toolset that helps industry to rapidly assess and utilize these new technologies. The optimization tools identify optimal equipment configurations and operating conditions for potential CO2 capture processes, thereby significantly reducing cost, time and risk involved in the implementation. The CCSI research group has developed two advanced optimization capabilities as part of its Framework for Optimization and Quantification of Uncertainty and Surrogates (FOQUS) tool. Both utilize GAMS as an essential element. The first tool performs simultaneous process optimization and heat integration based on rigorous models. The heat integration subproblem is modeled in GAMS as LPs and MIPs and solved by CPLEX. The other tool optimizes the design and operation of a CO2 capture system. The carbon capture system is represented as a MINLP model, which is implemented in GAMS and solved by DICOPT or BARON. GAMS is proud to be a part of this optimization toolset designed to make carbon capture a success.
www.gams.com
GAMS Integrated Developer Environment for editing, debugging, solving models, and viewing data.