Analytics November/December 2016 by Lionheart Publishing Inc.

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

NOVEM BER/ DE C EM BER 2016 BROUGHT TO YOU BY:

Crowdsourcing The power is in the numbers, and many hands make light work ALSO INSIDE: • Russia vs. Western analysts • Practical project management • The biggest big data challenges • Decision analysis software survey • Sports business analytics process

Executive Edge Verisk Chairman & CEO Scott G. Stephenson on the keys to unlocking an organization’s innovation

INS IDE STO RY

The wisdom of crowds As I write this, the 2016 U.S. presidential campaign staggers toward the finish line, leaving behind a trail of mud the likes of which we’ve never seen before. When the election is finally over, no matter the outcome, I think we all could use a hot shower. Like many of us, Analyze This! columnist Vijay Mehrotra craved a truly fair and balanced source of information to make sense of the election mess, particularly the parade of everchanging and often contradictory polls. For insight he turned to fivethirtyeight. com, the website created by Nate Silver that “uses statistical analysis – hard numbers – to tell compelling stories about elections, politics, sports, science, economics” and other topics. In this month’s column, “A ‘Silver’ lining for election blues,” Vijay takes a closer look at fivethirtyeight.com and how Silver and his team go to great lengths to make their forecasting models as accurate as possible, while acknowledging the limits of predictions. Telling an unbiased story, Vijay notes, isn’t easy, whether the topic is predicting politics or predictive policing. When it comes to predictions, I believe in the wisdom of crowds, especially

A N A LY T I C S - M A G A Z I N E . O R G

when people have to put their money where their mouths are. When there’s hundreds of thousands of people betting their own hard-earned money on the outcome of a U.S. presidential election or the World Series, you get a pretty good line on who’s the real favorite to win. And if you want in on the action, you better check your biases at the door. Crowdsourcing is a close cousin of the wisdom of crowds. Again, the idea is to tap into a worldwide source of opinions via the Internet, only this time the objective is to evaluate or address a particular task. As Ben Christensen, the director of content relevance operations at Appen, says in this month’s cover story on crowdsourcing, “The power of crowdsourcing is in its numbers.” Getting back to politics, which are hard to get away from these days, frequent contributor Doug Samuelson takes a look at how Russia’s intent on expanding its presence and influence necessitates new approaches to assessment by Western analysts (page 28). Now there’s a policy topic worth debating. Alas, like so much else during this presidential campaign, it was buried in mud. ❙

W W W. I N F O R M S . O R G

Want to Get Started with Prescriptive Analytics? Join a unique Collaborative Workshop and build an insightful prototype in 2 days!

Got a critical business challenge that you’d like to tackle with support from seasoned Analytics experts? Sign up for an AIMMS Proof of Concept workshop and experience the benefits of Prescriptive Analytics first-hand. Our POC workshop is a proven methodology that helps you and your team discover the value of Prescriptive Analytics and quickly develop a prototype to address a key business problem.

Get new business insights and perform rapid scenario analysis within 2 days

ty ili ib as Fe

Functional design What functionality do I want? Technical design How will it work? (data sources, integration)

2-8

Signoff

•

Conceptual design What do I need?

t Tes

design

Bridge the gap between business and technology

hn ical

•

Deploy

Te c

Engage the right stakeholders from the start

POC 2 days

•

Func tio na ld

Get the benefit of seasoned analytics professionals teaming with you to solve your toughest problems

Kick off

Requirements & Objectives

n ig es

•

Con cep tua ld es ig

Why join?

Full model build

Visit http://events.aimms.com/poc-workshop/ for more information Or contact us at info@aimms.com / 425-458-4024 What customers have to say: “The workshop far exceeded our expectations. The iterative process of the program was really amazing. I would have been happy coming out of this with 1 prototype and we got 9! It’s a Win-Win for everybody.” - Christian Olsen, Head of Business Development at Höegh Autoliners

C O N T E N T S

DRIVING BETTER BUSINESS DECISIONS

NOVEMBER/DECEMBER 2016 Brought to you by

FEATURES 28

PUTIN VS. WESTERN ANALYSTS Russia’s new approach to extending its influence necessitates new approaches to assessment. By Douglas Samuelson

PRACTICAL PROJECT MANAGEMENT Making analytics work: Why consistently delivering value requires effective project management. By Erick Wikum

CROWDSOURCING: CURATED VS. UNKNOWN Using the crowd: When to use a traditional crowdsourcing model and when to use an alternative. By Ben Christensen

DECISION ANALYSIS SOFTWARE SURVEY Past, present and future of dynamic software emphasizes continuous improvement of analytics tool. By Samantha Oleson

BIGGEST BIG DATA CHALLENGES Why half of big data projects fail: Don’t overlook the key role data lakes play in successful big data strategy. By Prashant Tyagi and Haluk Demirkan

THE SPORT BUSINESS ANALYTICS PROCESS Book excerpt: “Sport Business Analytics: Using Data to Increase Revenue and Improve Operational Efficiency.” By C. Keith Harrison and Scott Bukstein

64 4

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

Welcome to Analytic Solver ® Cloud-based Optimization Modeling that Integrates with Excel

Everything in Predictive and Prescriptive Analytics Everywhere You Want, from Concept to Deployment. The Analytic Solver® suite makes the world’s best optimization software available in your web browser (cloud-based software as a service), and in Microsoft Excel. And you can easily create models in our RASON® language for your server, web and mobile apps.

Linear Programming to Stochastic Optimization. It’s all point-and-click: Fast, large-scale linear, quadratic and mixed-integer programming, conic, nonlinear, non-smooth and global optimization. Easily incorporate uncertainty and solve with simulation optimization, stochastic programming, and robust optimization.

And it’s a full-power, point-and-click tool for forecasting, data mining and text mining, from time series methods to classification and regression trees, neural networks and association rules – complete with access to SQL databases, Apache Spark Big Data clusters, and text document folders.

Find Out More, Start Your Free Trial Now. In your browser, in Excel, or in Visual Studio, Analytic Solver comes with everything you need: Wizards, Help, User Guides, 90 examples, even online training courses. Visit www.solver.com to learn more or ask questions, and visit analyticsolver.com to register and start a free trial – in the cloud, on your desktop, or both!

Monte Carlo Simulation, Data and Text Mining. Analytic Solver is also a full-power, point-and-click tool for Monte Carlo simulation and decision analysis, with a Distribution Wizard, 50 distributions, 50 statistics and risk measures, and a wide array of charts and graphs.

Tel 775 831 0300 • Fax 775 831 0314 • info@solver.com

DRIVING BETTER BUSINESS DECISIONS

DEPARTMENTS

2 8 14 18 22 26 70 72 78

Inside Story Executive Edge Analyze This! Healthcare Analytics News & Notes Forum Conference Preview Five-Minute Analyst Thinking Analytically

Analytics (ISSN 1938-1697) is published six times a year by the Institute for Operations Research and the Management Sciences (INFORMS), the largest membership society in the world dedicated to the analytics profession. For a free subscription, register at http://analytics.informs.org. Address other correspondence to the editor, Peter Horner, peter.horner@mail.informs.org. The opinions expressed in Analytics are those of the authors, and do not necessarily reflect the opinions of INFORMS, its officers, Lionheart Publishing Inc. or the editorial staff of Analytics. Analytics copyright ©2016 by the Institute for Operations Research and the Management Sciences. All rights reserved.

A N A LY T I C S - M AGA Z I N E . O RG

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 at Austin Vice President Sections and Societies Esma Gel, Arizona State University Vice President Information Technology Marco Lübbecke, RWTH Aachen University Vice President-Practice Activities Jonathan Owen, CAP, General Motors Vice President-International Activities Grace Lin, Institute for Information Industry Vice President-Membership and Professional Recognition Susan E. Martonosi, Harvey Mudd College Vice President-Education Jill Hardin Wilson, Northwestern University Vice President-Marketing, Communications and Outreach Laura Albert McLay, University of Wisconsin-Madison Vice President-Chapters/Fora Michael Johnson, University of Massachusetts-Boston INFORMS OFFICES www.informs.org • Tel: 1-800-4INFORMS Executive Director Melissa Moore Director, Public Relations & Marketing Jeffery M. Cohen Headquarters INFORMS (Maryland) 5521 Research Park Drive, Suite 200 Catonsville, MD 21228 Tel.: 443.757.3500 E-mail: informs@informs.org ANALYTICS EDITORIAL AND ADVERTISING

Lionheart Publishing Inc., 506 Roswell Street, Suite 220, Marietta, GA 30060 USA Tel.: 770.431.0867 • Fax: 770.432.6969

President & Advertising Sales John Llewellyn john.llewellyn@mail.informs.org Tel.: 770.431.0867, ext. 209 Editor Peter R. Horner peter.horner@mail.informs.org Tel.: 770.587.3172 Assistant Editor Donna Brooks donna.brooks@mail.informs.org Art Director Alan Brubaker alan.brubaker@mail.informs.org Tel.: 770.431.0867, ext. 218 Advertising Sales Aileen Kronke aileen@lionhrtpub.com Tel.: 770.431.0867, ext. 212

Welcome to Analytic Solver ® Cloud-based Simulation Modeling that Integrates with Excel

Everything in Predictive and Prescriptive Analytics Everywhere You Want, from Concept to Deployment. The Analytic Solver® suite makes the fastest Monte Carlo simulation and risk analysis software available in your web browser (cloud-based software as a service), and in Microsoft Excel. And you can easily create models in our RASON® language for server, web and mobile apps.

Comprehensive Risk and Decision Analysis Tools. Use a point-and-click Distribution Wizard, 50 probability distributions, automatic distribution fitting, compound distributions, rank-order correlation and three types of copulas; 50 statistics, risk measures and Six Sigma functions; easy multiple parameterized simulations, decision trees, and a wide array of charts and graphs.

nonlinear optimization, simulation optimization, stochastic programming and robust optimization. And it’s a full-power tool for forecasting, data mining and text mining, from time series methods to classification and regression trees, neural networks and more, with access to SQL databases and Spark Big Data clusters.

Optimization, Forecasting, Data and Text Mining. Analytic Solver is also a full-power, point-and-click tool for conventional and stochastic optimization, with powerful linear and mixed-integer programming,

Tel 775 831 0300 • Fax 775 831 0314 • info@solver.com

EXE CU TIVE E D G E

Keys to unlocking innovation in organizations How to take a concept through infancy into reality and then on to profitable innovation.

BY SCOTT G. STEPHENSON

“Drive thy business or it will drive thee.” Benjamin Franklin offered this sage advice in the 18th century, but he left one key question unanswered: How? How do you successfully drive a business? More specifically, how do you develop the business strategy drivers that incite a business to grow and thrive? The 21st-century solution has proven to be data and analytics – from which emerge ideas large and small that can be the springboard to success. That solution begs yet another question: Where does an idea begin? And how can a concept be nursed through a promising infancy into reality and then guided to adulthood as a profitable innovation? As a leading provider of data and analytics, we at Verisk Analytics spend our time developing answer keys to such questions – keys that unlock innovation within an organization. THE C-SUITE VISION Lasting innovations often take root from evolutions in thinking. Data analytics is not a program or bolt-on

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

Welcome to Analytic Solver ® Cloud-based Data and Text Mining that Integrates with Excel

Everything in Predictive and Prescriptive Analytics Everywhere You Want, from Concept to Deployment. The Analytic Solver® suite makes powerful forecasting, data mining and text mining software available in your web browser (cloud-based software as a service), and in Microsoft Excel. And you can easily create models in our RASON® language for server, web and mobile apps.

Full-Power Data Mining and Predictive Analytics. It’s all point-and-click: Text mining, latent semantic analysis, feature selection, principal components and clustering; exponential smoothing and ARIMA for forecasting; multiple regression, logistic regression, k-nearest neighbors, discriminant analysis, naïve Bayes, and ensembles of trees and neural networks for prediction; and association rules for affinity analysis.

distributions, 50 statistics and risk measures, rankorder and copula correlation, distribution fitting, and charts and graphs. And it has full-power, point-and-click optimization, with large-scale linear and mixed-integer programming, nonlinear and simulation optimization, stochastic programming and robust optimization.

Simulation/Risk Analysis, Powerful Optimization. Analytic Solver is also a full-power, point-and-click tool for Monte Carlo simulation and risk analysis, with 50

Tel 775 831 0300 • Fax 775 831 0314 • info@solver.com

EXE CU TIVE E D G E

accessory for business operations; it’s a strategy that should be embraced as sinew and fiber through the body of an organization. Instead of an afterthought, analytics must stand at the forefront of business strategy and be embraced throughout the organization – and not just by a data analytic “elite.” A business so tooled for the 21st century will seamlessly use data from within and outside the enterprise to generate new and unique insights. To realize this vision, corporate leadership should ask: Are line managers thinking, behaving and sounding more like data analysts? And are data scientists and data engineers sounding more like line managers? Ideally, both the business mindset and the data analytic mindset should reside between the ears of the same individual. But even short of the ideal, the goal is to assimilate the notion of using data to create meaningful information and not merely to warehouse it. Companies need – and can achieve – continuous, agile improvements propelled by the skillful use of data analytics. If C-suite executives can harness the power and fluidity of data analytic methods, they should insist that projects deliver results within nine months. A monolithic plan or “Manhattan Project” isn’t necessary to derive substantial benefits. 10

A N A LY T I C S - M A G A Z I N E . O R G

Looking ahead, risk decisions will become more targeted and in real time with the advent of remote imagery, textual analyses, connected homes and the Internet of Things. Analysts estimate that current North American smart-home penetration may reach 25 percent to 30 percent by 2020. What this all means is that new technology is providing us with different kinds of data points, in greater volumes and with faster speed. That information will enhance analytic models to improve risk selection and assessment. THE CHOICE TO BE INNOVATIVE Two things about innovation are true: It’s sometimes expensive, and it’s never without risk. That’s the trade-off. And because of the trade-off, a company has to make a conscious decision to become a leader in innovation. That said, if you look across any industry in the world with a technological basis, the reward for being innovative, and the penalty for not, are both greater than they ever were. In almost any industry, the differential in performance based on innovation has spread – it’s bigger than it used to be. In the pursuit of innovation, businesses first need to determine what innovation means for them. Will the innovation address process or product, service, pricing or any number of value propositions? How deeply W W W. I N F O R M S . O R G

entrenched will innovation be in the organization’s thought processes? How much will the organization commit? How will the business demonstrate an innovative process or product as one of its distinguishing characteristics as it goes to market? The next choice is how to sequence the innovations. There’s no single correct path. What needs to be done first or third may be based on what’s important to the customer or simply what appears to be the logical progression – that is, what needs to be built first, which is then built on and built on again.

Ivan B. Class of ‘18 Oil and Gas

For every company, the sequence will be different. And that’s what makes innovation interesting. Those are the decisions we make at Verisk every day. DATA’S COMPETITIVE EDGE Today, to be competitive, companies must strive for something Verisk has dubbed “the n+1 data set.” If a company’s data set has a certain number of elements (n), that set should be stretched to include one more. Elements must continually be added, advancing toward the next layer and adding richness to the analysis.

“This program has helped me develop some great tools for my analytics belt.” -Spring 2016 Exit Survey

MORE INFORMATION: ANALYTICS.STAT.TAMU.EDU

Informs ad.indd 1

A NA L Y T I C S

10/17/16 2:57 PM

N O V E M B E R / D E C E M B E R 2 016

EXE CU TIVE E D G E

Although this process requires investment in data resources, analytics, technology and people, the return on investment can mean thriving, rather than merely surviving or even failing. While it’s true for all industries, it’s especially relevant in data-driven industries such as insurance, energy and financial services. Let’s consider two examples of how to apply n+1: Example 1: In the traditional approach, fraud investigators start with single-suspect data points, such as a name or address. Then they build a network with the associated data. However, a newer, more proactive approach scans data sets to detect fraudulent networks, uses advanced analytics to identify network attributes, and then scores and prioritizes those networks based on their fraud potential. Finally, when we overlay data from social media or data derived from unstructured data, such as mining the text of claims adjusters’ notes, and then apply the new data-enhanced analytics, we have a more comprehensive toolbox to use in fighting fraud. Example 2: Relentless pressure to lower coal consumption will intensify competition among producers for the remaining

coal market. Some producers will meet this challenge using a conservative approach and focus on mining coal, which is what they do best. As market opportunities fade, only low-cost suppliers with access to primequality coal and superior market knowledge will succeed. Ultimately, it will become a survival-of-the-fittest contest. However, diversification is another strategy, and that would rely on the latest data and analytics. Diversification for coal producers is a “stretch” strategy aimed at participating in a wider, non-coal energy market that’s growing, not declining. So again, here we clearly see the value of the n+1 mindset. At Verisk, the seed of innovation began with a question that prompted more questions and created a company culture that finds solutions. As Ben Franklin also once said, “An investment in knowledge always pays the best interest.” ❙ Scott G. Stephenson is chairman and chief executive officer of Verisk Analytics. Verisk’s mission is to help customers understand and manage the risks they face every day. On Oct. 7, 2009, Verisk Analytics debuted on the NASDAQ Global Select Market as the largest domestic IPO of the year. Four decades of continuous innovation were recognized in 2015 and 2016 when Verisk Analytics was named to the Forbes list of the World’s Most Innovative Companies.

Request a no-obligation INFORMS Member Benefits Packet For more information, visit: http://www.informs.org/Membership

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

CPLEX Optimization Studio®. Still the best optimizer and modeler for the transport 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 transport industry for developing and deploying models and optimizing transport strategy. 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.

ANALY ZE TH I S !

A ‘Silver’ lining for election blues For analytics professionals, the world of presidential elections is familiar territory.

BY VIJAY MEHROTRA

For the past several months, I have spent hours staring at my screen, reading anything I can get my hands on that might help me get a sense of what might happen during the elections on Nov. 8. Since I live in Oakland, Calif., the heart of the uber-liberal bubble that is the San Francisco Bay Area, I am constantly searching for truly fair and balanced perspectives about what is really going on across the rest of the country, especially with regards to this year’s presidential election. For analytics professionals, in fact, the world of presidential elections is familiar territory. Since George Gallup first applied statistical random sampling to draw conclusions about the results of the 1936 U.S. presidential election based on survey data, there has been a steady (and recently explosive) growth in the number of people gathering, analyzing, visualizing and interpreting data to try to understand, explain, predict and/or influence what might happen in our elections. Nowadays, during our seemingly endless presidential election campaigns, it feels as though there are new state or national poll results being announced every day for months. This almost nonstop and often contradictory stream of numbers often leaves me bewildered.

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

Which is why I spend a huge amount of time visiting fivethirtyeight.com, the political website created some years ago by Nate Silver, a “number-crunching prodigy” [1] who just might be the big data world’s first mainstream rock star. Since bursting onto the political scene in 2008 with predictions that: (a) were somewhat contrary to the punditry’s consensus and (b) often proved to be surprisingly accurate, Silver’s website has sought to use both historical voter data and information from political polls to make predictions about elections (the site, now owned by ESPN, also includes data-driven stories about sports, science, economics and culture). During this presidential election cycle, Silver and his team used one set of models for forecasting state-by-state primary results and another set for the general election. Quite a bit of methodological detail about these forecasts is publicly available at fivethirtyeight.com [2,3]. After reading Silver’s book “The Signal and the Noise” [4] and scrutinizing these forecasting process descriptions, my sense is that Silver and his team are seeking to understand the same elusive truths about the electorate that I am, and as an engaged citizen I am grateful for their efforts. The blizzard of polling data is systematically examined, with some polls banned due to ethical and/or methodological concerns. Careful

weighting is done to account for factors such as sample sizes, recency, frequency and past performance. A wide variety of adjustments are made to address factors such as post-convention bounces, third-party candidates, registered vs. likely voters and historical biases associated with particular polls. Moreover, over the past eight years Silver and his team have continued to tweak their models in different ways, and there appears to be a basic humility about the limits of prediction underlying their observations and claims, as well as a deep-seated wonky desire to tell an unbiased story. Which brings us back to Oakland. In a recent article in Significance [5], Kristian Lum (lead statistician at the Human Rights Data Analysis Group) and William Isaac (doctoral candidate in the Department of Political Science at Michigan State University) examine the controversial topic of “predictive policing [6],” a term that refers to the use of data and models to make forecasts about where crime is most likely to take place, usually within urban population centers. While predictive policing software has been commercially available for some time, its efficacy has been hotly debated in crimefighting circles [7]. Lum and Isaac’s main thesis is that the historical data used by these predicting policing systems is

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

ANALY ZE TH I S !

inherently biased, and that this bias is in turn propagated by the machinelearning algorithms that are embedded in the software. Citing a great deal of previous research, these authors are inherently suspicious about this historical data, concluding that “police records do not measure crime. They measure some complex interaction between criminality, policing strategy, and community-police relations.” Using data about Oakland, the authors tell a thoughtful and illuminating data-driven story. Rather than accepting police data as their proxy for actual drug crime, they instead use data from the 2011 National Survey on Drug Use and Health (NSDUH). After making a solid case for why the NSDUH data is likely to provide a more accurate snapshot of drug use in Oakland than past arrest data, the paper then contrasts the NSDUH data with the actual police arrest data for drug crimes for 2010, pointing out that “while drug crimes exist everywhere, drug arrests tend to only occur in very specific locations – the police data appear to disproportionately represent crimes committed in areas with higher populations of non-white and low-income residents.” The authors then make an important observation about how biases can quickly propagate even more 16

A N A LY T I C S - M A G A Z I N E . O R G

quickly than the machine-learning models would suggest: “But what if police officers have incentives to increase their productivity as a result of either internal or external demands? If true, they might seek additional opportunities to make arrests during patrols. It is then plausible that the more time police spend in a location, the more crime they will find in that location.” Putting all of this together, the overarching premise here is as follows: (a) algorithms based on biased input data suggest that crime will be found in certain areas, which leads to (b) more policing in those areas, which causes (c) dramatically larger numbers of arrests in those areas relative to other less policed areas, which leads to (d) increased bias in the input data for the algorithms. The paper concludes by presenting the results of a simulation that vividly illustrates this insidious cycle. In the big data age, our understanding of the world and the future – and our decisions about what actions to take based on that understanding – are increasingly dependent on algorithms. As analytics professionals, most of us have some inherent bias toward data-driven methods, but often this bias should be tempered with a healthy dose of skepticism W W W. I N F O R M S . O R G

about such models and about the data that drives them. Political polling and predictive policing are just two pernicious examples of how biased data can distort our beliefs and behaviors – but as a darkskinned foreigner living in America during this year’s presidential election and a proud resident of Oakland, they hit really particularly close to home for me. ❙

REFERENCES & NOTES 1. http://nymag.com/news/features/51170/ 2. http://fivethirtyeight.com/features/how-weare-forecasting-the-2016-presidential-primaryelection/ 3. http://fivethirtyeight.com/features/a-usersguide-to-fivethirtyeights-2016-general-electionforecast/ 4. https://www.amazon.com/Signal-Noise-ManyPredictions-Fail-but/dp/0143125087 5. Significance is joint publication of the Royal Statistical Society and the American Statistical Association, available online at www.rss.org.uk/ significance 6. http://onlinelibrary.wiley.com/doi/10.1111/j.17409713.2016.00960.x/full

Vijay Mehrotra (vmehrotra@usfca.edu) is a professor in the Department of Business Analytics and Information Systems at the University of San Francisco’s School of Management and a longtime member of INFORMS.

7. See for example http://www.sciencemag.org/ news/2016/09/can-predictive-policing-preventcrime-it-happens

Which organization has the best O.R. department in the world?

prize

call for

nominations

DEADLINE TO APPLY IS DECEMBER 1, 2016

The Institute for Operations Research and the Management Sciences annually awards the INFORMS Prize for effective integration of Operations Research/Management Science (OR/MS) and advanced analytics into organizational decision making. The award is given to an organization that has repeatedly applied the principles of OR/MS and advanced analytics in pioneering, varied, novel, and lasting ways.

log onto: www.informs.org/informsprize for more information or contact Tarun Mohan Lal, Principal, Advanced Analytics Clinic Operations Mayo Clinic voice: +1 507-266-9842 email: mohanlal.tarun@mayo.edu

2017

Tell us why the title should be yours.

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

HEALT H CARE A N A LY T I C S

Data governance and management The fundamental building blocks for a sustainable data analytics program

Healthcare organizations jumped on the data analytics bandwagon without establishing a process of data governance.

BY RAJIB GHOSH

The healthcare analytics industry is making great strides. As part of my work I talk to many data analytics companies who report they are very busy with implementation projects. One large company told me that its implementation staffs are booked until end of the first quarter of 2017. This is evidence that the demand for healthcare analytics is strong. I predicted the rise in demand during the latter part of 2016 in my previous articles. Payment reform has started to accelerate, with more states and commercial payers moving their contracts with providers from volume to value. Without a strong data analytics program, it is impossible for many provider organizations to stay viable in this new environment. Many data analytics companies report their clients are missing the basic building blocks for a successful data analytics program: data governance and data management. Healthcare organizations

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

jumped on the data analytics bandwagon without establishing a process of data governance. In this article I will outline the need for data governance and my experience leading such an initiative within a complex organization. DATA HAS A TIME VALUE

Figure 1: The longer it takes to collect data and complete analysis, the less effective it becomes in informing decisions.

Data has a time value, and this is not a secret. Information and insights from data have the best value when the business is attempting to understand why something just happened. The longer the process takes, the less this value becomes. For data analytics to produce the best return on investment, it is important for the business to have the relevant data available at the fingertips of the analyst so that the analysis can be produced quickly to help decision-makers make decisions in a timely fashion. Therefore, the longer it takes to collect all the data and conduct the analysis the less effective the action becomes for the business (see Figure 1). According to Health Catalyst, a leading U.S. healthcare analytics company, about 80 percent of an analystâ&#x20AC;&#x2122;s time is spent in searching for the relevant data. That is a huge waste of expensive resources and valuable time for any organization.

DATA QUALITY: THE OTHER BIG ISSUE Access to data alone is not enough. Access to good quality data is extremely important to make the work of analytics worthwhile. According to IBM, one in three business leaders today in the United States do not trust their own data. Poor data quality costs the U.S. economy around $3.1 trillion a year. What is the point in conducting sophisticated analysis with expensive data analysts and scientists if that is the case? In my experience data quality analysis is not always done adequately within healthcare organizations. It is one of the most tedious tasks that seldom bears any glamor. It is, however, a key function of what the data industry calls the role of a data steward. A data steward needs to have the necessary knowledge about the content and the metadata to assess

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

Source: Rajib Ghosh

HEALT H CARE A N A LY T I C S

the quality of the data, and then work with business units to correct what is wrong. Sometimes it may lead to fixing issues with the data acquisition process and even standardizing the data vocabulary. DATA GOVERNANCE: KEY PIECE OF THE PUZZLE Timely access and quality bring us to data governance. Organizations need to invest time and resources to build a well-defined data governance process. It begins with identifying key people to participate in an organizational data governance committee. Many times the committee can be appointed by the CEO or the board of directors to ensure that the data in the organization is treated like an asset. In the case of healthcare organizations, members of such a committee may include the chief data officer, chief analytics officer, chief financial officer, chief medical officer, chief information officer, chief operating officer, etc. In other words, fairly senior members of the organization or top business unit leaders. A data governance committee has to ensure that the organizationâ&#x20AC;&#x2122;s data is governed appropriately, maintenance of metadata definitions and business rules are followed, and appropriate levels of data privacy and security audits are in 20

A N A LY T I C S - M A G A Z I N E . O R G

place. Healthcare organizations benefit when they bring their clinical, operational and financial data together to develop a single definition of truth. To achieve that goal, the data governance committee needs leadership representation from all those functional areas. DATA GOVERNANCE FOR A NETWORKED ORGANIZATION I was asked to chair a data governance committee of a network of several healthcare organizations. The goal was to build data governance policies and procedures for building a centralized data analytics infrastructure. To achieve this goal we established a committee with leadership representation from all the functional areas as stated in the previous section. We embraced a framework that focused on four key areas: governance, stewardship, management and compliance. We diligently worked on various policies and procedures that not only addressed the needs of the present day but also considered emerging opportunities such as data on social determinants of health, mental health and substance use. To establish transparency in data flow between networked organizations, we implemented strict data access monitoring and reporting policies and procedures. We also defined oversight of data W W W. I N F O R M S . O R G

stewardship as a key role of the committee. Committee members were given the responsibility of developing data for a stewardship program within their own organization as well as within the centralized data organization that assumed the responsibility for the data analytics program. The committee oversaw specifications of data extraction, transformation and load specifica- Data quality analysis is not always done adequately within healthcare organizations. Photo Courtesy of 123rf.com | scanrail tions, along with adequate data security and privacy measures. The governance is in place, the data team upfront time spent in crafting the goverof any organization can feel confident nance process enabled this complex netthat they have established a sustainable work of organizations to develop a data and effective analytics program that analytics infrastructure with confidence will eventually garner kudos in the and transparency. boardrooms. â?&#x2122; Data governance is a fundamental Rajib Ghosh (rghosh@hotmail.com) is an building block for successful and independent consultant and business advisor with 20 sustainable data analytics programs for years of technology experience in various industry verticals where he had senior-level management organizations of any size and complexity. roles in software engineering, program management, It is not a glamorous job. It does not product management and business and strategy development. Ghosh spent a decade in the U.S. create press cycles. It is also not very well healthcare industry as part of a global ecosystem understood by executives responsible for of medical device manufacturers, medical software companies and telehealth and telemedicine solution data infrastructure. However, technology providers. Heâ&#x20AC;&#x2122;s held senior positions at Hill-Rom, for data analytics has now become a Solta Medical and Bosch Healthcare. His recent work interest includes public health and the field of commodity with many vendors striving IT-enabled sustainable healthcare delivery in the to earn enterprise business. Once the United States as well as emerging nations. fundamental building block of data A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

NE W S & N OT E S

Salary survey, O.R. degree, CAP updates Women receive 95+ percent of their male counterparts’ salaries over all six categories studied.

SALARY SURVEY: PREDICTIVE ANALYTICS PROFESSIONALS

The gender pay gap among predictive analytics professionals is practically nonexistent, and the percentage of U.S. citizens in the profession at the junior level is increasing while the percentage of foreign-born professionals at the same level is decreasing as more U.S. students are entering the market. Those are two of the more notable trends in a new Burtch Works salary survey of predictive analytics professionals (PAPs). According to the 2016 survey data, women earn the same or slightly higher base salaries than men in two of the six job categories studied, and receive 95+ percent of their male counterparts’ salaries over all six categories. The study gathered compensation and demographic data from 1,216 PAPs representing more than 650 companies nationwide. For the purpose of the study, a PAP was defined as those who can “apply sophisticated quantitative skills to data describing transactions, interactions or other behaviors of people to derive insights and prescribe actions. PAPs Gender pay gap among predictive analytics professionals is next to nil. Photo Courtesy of 123rf.com | zozoen are distinguished from business 22

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

intelligence professionals or financial analysts by the enormous quantity of data with which they work, well beyond what can be managed in Excel. However, this definition also encompasses data scientists, who are excluded from this study because of their distinguishing ability to work with unstructured data, resulting in different compensation.” PAPs’ median base salaries range from $75,000 for Level 1 individual contributors to $225,000 for Level 3 managers. Nearly three-quarters of all PAPs are eligible for bonus pay, which can substantially increase their income. Education, experience, skill set, industry and region all impact compensation. For example, at almost all job levels, PAPs employed on the West Coast and in the Northeast earn the highest salaries. Founded by Linda Burtch, Burtch Works Executive Recruiting is a leading resource for highly qualified analytic and marketing research talent in the United States. To download a free copy of the complete Burtch Works Study, “Salaries of Predictive Analytics Professionals,” click here.

Individuals with a master’s degree in operations research rank third highest in terms of mid-career salaries, according to Forbes. Photo Courtesy of 123rf.com | dotshock

Individuals with a master’s degree in operations research rank third highest in

terms of mid-career salaries, according to a recent survey by Forbes. The findings, part of Forbes’ 2016-2017 College Salary Report, combed data for nearly 200 graduate-level degrees and ranked them according to the mid-career salaries of full-time U.S. employees who hold specific degrees. Mid-career operations researchers with master’s degrees average $129,000 in annual salary, according to the survey, just below those with master’s degrees in nurse anesthesia ($156,000) and computer science and engineering ($134,000) and just ahead of those with master’s degrees in electrical and electronics engineering and taxation. MBAs were omitted from the survey. As expected, those with master’s degrees in STEM disciplines dominated the top of the list, with chemical engineering, biomedical engineering, industrial engineering and applied mathematics all finishing in the top

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

FORBES SURVEY: MASTER’S DEGREE IN O.R. PAYS OFF

NE W S & N OT E S

10 with mid-career wages between $115,000 and $121,000. For more information, click here. ‘GUIDE TO THE ANALYTICS BODY OF KNOWLEDGE’ The INFORMS “Guide to the Analytics Body of Knowledge” (ABOK) is nearing a first draft. Along with a study guide for those seeking Certified Analytics Professional (CAP®) status, ABOK can also be used by instructors to help with curriculum development, by HR managers to determine what the employees need to know, by regulators to understand the state of analytics practice, by employers to provide a career path for employees and by newcomers to analytics to understand more about the field. A proposed table of contents includes chapters on “Introduction to Analytics,” “Getting Started with Analytics,” “The Analytics Team,” “The Data,” “Solution Methodology,” “Model Building” and “Deployment and Lifecycle Management.” ABOK Committee Chair Terry Harrison (Penn State University) and Editor in Chief James Cochran (University of Alabama) will lead a session on ABOK on Nov. 15 in Nashville as part of the 2016 INFORMS Annual Meeting. Conference attendees are invited to join the ABOK discussion session.

A N A LY T I C S - M A G A Z I N E . O R G

WILLIAMS TO LEAD 2017 ANALYTICS CERTIFICATION BOARD Jim Williams, CAP, of FICO (photo) and Aaron Burciaga, CAP, of Accenture Analytics, have been elected chair and vice chair, respectively, of the 2017 Analytics Certification Board (ACB). The ACB, Jim Williams whose members are elected by INFORMS and CAP® designees, meets on a quarterly basis to provide oversight and guidance to the Certified Analytics Professional (CAP) and the Associate Certified Analytics Professional (aCAP) programs. Randy Bartlett, CAP, Blue Sigma Analytics; Russell Barton, CAP, Penn State University; Tom Davenport, Babson College; Bill Franks, Teradata; Jeanne Harris, Columbia University of New York; Stefan Karisch, Boeing; Lisa Kart, CAP, Gartner; Jack Levis, UPS; Polly MitchellGuthrie, SAS; Jonathan Owen, CAP, GM; Michael Rappa, NCSU/Institute for Advanced Analytics; Greta Roberts, Talent Analytics; Nick Wzientek, CAP, Vistar; and Melissa Moore, INFORMS executive director, complete the board. Karisch (2017 president of the Analytics Society) and Barton (2017 INFORMS VP for Sections and Societies) will serve as ex-officio members. For information, click here. ❙

W W W. I N F O R M S . O R G

Stop compromising. With FICOâ&#x20AC;&#x2122;s advanced analytics platform, you can deploy your existing math models with what-if analysis and visualization in a matter of weeks. Supported by 60 years of proven industry experience, you can provide trusted insights to increase productivity and improve profitability.

FO RUM

A conceptual framework for BI/analytics strategies Based on research and best practices, the framework aims to bring clarity to the field.

BY JAY LIEBOWITZ

Business intelligence (BI) and analytics programs are among the “hottest” curricula being developed at universities and colleges worldwide. Definitions vary on what entails business intelligence or analytics, and there doesn’t seem to be a universal BI/analytics conceptual framework that is being used by organizations and universities to develop their BI/analytics strategy and associated roadmap. To provide some clarity and based on research and best practices in the field, I developed the BI/Analytics Conceptual Framework as shown in Figure 1. In order to better understand if these are the right components, I reached out to some BI/analytics experts in industry and universities as part of a Delphi survey. Some of the preliminary comments from those experts in this first round include: • Analytics skills mature over time within organizations, suggesting the value of incorporating a CMM (capability maturity model) in your framework. • Other business and IT drivers might include: different skill levels in working with voluminous data; visibility into competitors’ moves so competitive responses can be developed; being able to combine customer-provided data with other information we have about those same customers; curating and filtering information into

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

“need to know” slices so confidentiality and privacy are protected. • Other BI enablers may include: analytics to become trusted advisors to senior executives (this requires more than technical analytic skills – it requires deep understanding Figure 1: BI/analytics conceptual framework. Framework author Jay Liebowitz used a Delphi survey of experts for verification. The red items indicate the more important factors based on the of the business and surveyed experts. marketplace, strong influencing and relationship-building analytics processes; applicability skills, organizational savvy, effective of the results; connect key risk storytelling and visualization skills, indicators with key performance and a willingness to present candidly indicators. even unwelcomed information); organizational design can help The second round with the Delphi or hinder the impact of analytic experts identified the red highlighted investments; problem definition and factors in Figure 1 as being the most problem prioritization. important in the framework. Before going • Other BI/analytics strategy goals: ahead in further revising the conceptual reduce speculation and judgment framework, I would be curious in getting bias that affect objectivity and your feedback as to the accuracy and barriers imposed by “hidden” factors completeness of this proposed BI/ in the decision-making process. analytics conceptual framework. • Other BI/analytics success factors: I welcome input and thoughts. Send define problems correctly (digging comments to jliebowitz@harrisburgu.edu. ❙ and not just reviewing the surface); Jay Liebowitz (jliebowitz@harrisburgu.edu) is the preparedness for the analytics DiSanto Visiting Endowed Chair in Applied Business and Finance at Harrisburg University of Science and process (collaboration); management Technology in Harrisburg, Pa. He is a member of of expectations about outcomes of INFORMS.

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

STU DY: IN T E R N AT I O NA L I N T R I GU E

Krelim vs. Western analysts Russia’s new approach to extending its influence necessitates new approaches to assessment.

BY DOUGLAS A. SAMUELSON ussia has a plan to take over Central and Eastern Europe, only this time by buying it rather than overrunning it. Russia’s recent move to a more assertive foreign policy has more and more analysts trying to guess its intentions and how the Western world can respond. Russia’s military push into Georgia, the advance of rebels presumably backed by Russia in the Crimea, Russia’s possible involvement in the hacking of Democratic National Committee emails and generally bolder

A N A LY T I C S - M A G A Z I N E . O R G

statements in foreign policy indicate that there are reasons for concern. But there is broad disagreement about what Russia’s objectives and plans really are. A good outline, backed by substantial data, appears in a report recently completed by study teams at the Center for Strategic and International Studies (CSIS), one of Washington, D.C.’s most respected think tanks, and the Center for the Study of Democracy (CSD) in Sofia, Bulgaria [1]. As the study’s authors quoted Marti Dempsey, then chairman of the Joint Chiefs of Staff, W W W. I N F O R M S . O R G

Russian President Vladimir Putin’s playbook focuses on economic influence in Central and Eastern Europe. Photo Courtesy of 123rf.com | Igor Dolgov

from July 2014, “[Vladimir Putin] got a playbook that has worked for him now two or three times, and he will continue to [use it.]” The analysts focused on Russian economic influence in Central and Eastern Europe, with detailed attention to events from 2004 through 2014 in five countries: Hungary, Bulgaria, Latvia, Slovakia and Serbia. The study’s authors’ conclusion is that since 2008 or thereabouts, Russia has deliberately and systematically pursued increasing economic influence in the countries of Central and Eastern Europe, trying to draw them away from close cooperation with the United States and the European

Union (EU). These pursuits include substantial efforts to involve those countries’ leaders in financial arrangements to enrich the leaders in return for more pro-Russian and anti-democratic policies and politics. Detailed depiction of the five countries suggests some factors that affect resilience against Russia’s creeping influence.

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

DETERIORATION IN DEMOCRATIC GOVERNANCE The study’s authors call Hungary “an early adopter of illiberalism.” They wrote, “At the beginning of the study period, Hungary was among the best performers in Central Europe in terms |

RU S S IA VS . W E ST E R N A N A LYST S

of its good governance practices and was steadfast in its Euro-Atlantic orientation, having acceded to NATO in 1999 and the European Union in 2004. Yet, as the decade progressed, Hungary experienced a marked deterioration in democratic governance standards following its 2010 parliamentary elections.” The government has revised the constitution five times, increasingly restricted media and minorities, and imposed more control over the judiciary and the central bank. These developments have coincided with reduced cooperation with NATO, rising dependence on Russian energy supply and increasing new projects with Russia. Also during this time, the authors assert, the think tank Transparency International stated that Hungary has adopted “a centralized form of corruption that has been built up and made systematic.” In Bulgaria, the study’s authors find that economic linkages, particularly in energy, “provide the Kremlin with considerable leverage over current and future decision-making.” This happens “through an interplay of reinforcing networks of influence that range from corrupt politicians and like-minded political parties to energy majors and Bulgarian oligarchs.” Among the five study countries, Bulgaria has the largest proportion of Russian investment and the most pervasive influence on policy, including legal changes that make 30

A N A LY T I C S - M A G A Z I N E . O R G

financial dealings more opaque. This set of effects led the study’s authors to call Bulgaria an example of “state capture” by the Russians. Latvia, in contrast, offers a model of how to “break the unvirtuous cycle.” Despite – or perhaps because of – its close proximity to Russia and its former status as a Soviet republic, Latvia has resisted “the malign influence of the Russianlinked oligarchies” most effectively of the five study nations. The authors suggest that this may be because Latvia “has more successfully deepened its nascent democratic institutions and reinforced the rule of law” and the deep skepticism by most Latvians of Russian intentions. Their historical narrative is one of Soviet occupation rather than liberation from the Nazis, and this seems to have made them less persuadable by current Russian initiatives. Thus Latvia, despite the second highest proportion of Russian investment, seems most resilient against Russian influence. Slovakia, the study’s authors said, has seen a steady decline in democratic governance rankings over the past decade – less pronounced than in Hungary but persistent. While Slovakia did attempt to “dismantle communist-era networks and shrug off authoritarian and nationalistic-style leadership” between 1998 and 2006, “many Slovaks W W W. I N F O R M S . O R G

did not enjoy the economic benefits, which exposed the less firmly planted roots of Slovakia’s liberal, democratic tradition.” The ensuing government made financial transactions more opaque and introduced more policies and practices widely seen as unfair. The study’s authors conclude that Slovakia is an example of a “state capture in action.” Serbia, in contrast to Latvia, has deep historic ties to Russia that extend to the recent past, including Russian opposition to NATO intervention in the Balkan conflicts of the 1990s. Hence “the Kremlin has continued to deepen its bilateral ties to Serbia even as Serbia has sought to become a member of the European Union.” These activities include heavy Russian investments in key sectors of Serbia’s economy, such as railway equipment and infrastructure. Serbian companies and Serbian politicians, often thoroughly obscured by complex contractual relationships, seem to have greatly expanded Russia’s political influence in Serbia. The study’s authors rather gloomily describe Serbia as “a preview of coming attractions.”

To counter the tendencies they identified, the study’s authors made several policy recommendations:

• Elevate and design a specific, highlevel task force within the U.S. Financial Crimes Enforcement Network that focuses solely on tracing and prosecuting illicit Russian-linked financial flows if they interact with the U.S. financial system. • Encourage NATO and EU members to task their own financial intelligence units with developing dedicated units that track illicit Russian transactions. • Prioritize EU-U.S. financial intelligence cooperation. • Elevate anti-corruption by strengthening institutions as an element of NATO’s Readiness Action Plan. • Completely revamp U.S. assistance to Central and Eastern Europe and the western Balkans to prioritize combating Russian influence and strengthening governance. • Substantially enhance EU member states’ and institutions’ anti-corruption and development assistance mechanisms to help the most vulnerable countries build greater resilience to Russian influence. • Introduce more rigorous benchmarking of rule of law and anti-corruption efforts as-conditions for pre-accession aid to the western

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

POLICY RECOMMENDATIONS

RU S S IA VS . W E ST E R N A N A LYST S

Balkans and other countries seeking EU accession. • Earmark specific EU-wide and national funds for support of the rule of law. • Enhance EU oversight of EU development funds and require full disclosure of company ownership when meeting EU diversification requirements. ANALYTICS OPPORTUNITIES The study did leave open some interesting issues analytics professionals might wish to pursue. The definition of “corruption” is broad, not tied to an international legal standard, and could therefore be hard to distinguish from what the Western democracies might call “targeted incentives.” More precise terminology and metrics would sharpen the analysis, perhaps yielding better recommendations about how to counter the more pernicious forms of influence. The study relied heavily on rankings from Freedom House, a well-respected organization specializing in assessments of governmental institutions around the world, for measures of democratic government, judicial independence, and other aspects of economic and political “health.” Even the best such organizations can have biases and blind spots, however, especially when 32

A N A LY T I C S - M A G A Z I N E . O R G

relying on subject matter experts’ judgments. Comparison against one or more other sets of assessments would most likely be illuminating. Although the study’s authors described the pattern they observed as a “network flow model,” they did not cite a reference in either the operations research/optimization literature or the social networks literature. Neither did they discuss what cuts in the network would disable it. Input-output economics would also be illuminating. Here, too, some analytics attention could be helpful. The study focused entirely on actions within Eastern and Central Europe in the countries Russia attempts to influence. Analysis of the inside-Russia component of these activities was left for others to pursue. Hopefully, someone will. In particular, it would be interesting to assess whether Russian actions are as cohesive as they may appear from the vantage point of the countries studied, and whether their intent is entirely malevolent. The study does not mention wargaming, although some wargames have elucidated financial and other noneconomic actions as part of hypothesized conflicts between Russia and NATO (see, for instance, [2].) With development of better model-based assessments W W W. I N F O R M S . O R G

of consequences, such wargames could be quite valuable in suggesting Russian courses of action and Western countermeasures. The study does provide a serious, in-depth, careful look at the means of Russian influence, apparently directed at weakening Euro-Atlantic ties and hence undermining U.S. interests. These issues deserve much additional scrutiny both by analysts and policy-makers, as politicoeconomic influence appears to be a much more consequential and imminent threat to the United States, both the nation and

its businesses, than the still-daunting Russian nuclear arsenal. ❙ Douglas A. Samuelson, D.Sc. in operations research, is president and chief scientist of InfoLogix, Inc., a small R&D and consulting company in Annandale, Va. He is a longtime member of INFORMS. REFERENCES 1. Heather A. Conley, James Mina, Ruslan Stefanov and Martin Vladimirov, October 2016, “The Kremlin Playbook: Understanding Russian Influence in Central and Eastern Europe,” Rowman and Littlefield, Lanham, Md. Also available as a pdf download online from www.csis.org. 2. Douglas A. Samuelson and Russell R. Vane III, June 2015, “Wargamers Explore ‘Forbidden Options,’ ” OR/MS Today.

DAIL S W eNE ADVERTISE IN ANNUAL MEETING eNEWS DAILY

Reserve your advertising space now in the Annual Meeting eNews Daily, the official news source of the INFORMS Annual Meeting. Each morning attendees will receive the Annual Meeting eNews Daily in their inboxes. eNews Daily provides a preview of the day’s key events and a recap of the previous day’s happenings. Over 60% of the 5,000+ attendees open and read eNews Daily.

DELIVERY SCHEDULE:

eNews Daily will be delivered by 5am on the following days:

ADVERTISING RATES: Any 1 day................ $250 Any 2 days.............. $450

Wed., 11/9............... preview edition to all meeting pre-registrants

Any 3 days.............. $600

Sun., 11/13.............. all attendees

All 5 days................ $750

Mon., 11/14............. all attendees

Any 4 days............... $700

Tues., 11/15............. all attendees Wed., 11/16............. all attendees

QUESTIONS:

Nashville 2016

Please direct questions to Tom Fink, Senior Marketing Manager Phone: 443-757-3561, email: tom.fink@informs.org

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

AN ALY TIC S M AT U R I T Y

Making analytics work through practical project management Projects are the primary vehicles through which we deliver value, and consistently delivering value with analytics requires effective project management.

BY ERICK WIKUM here does your organization fit into the project management spectrum shown in Figure 1? On one end of the spectrum, labeled “Do Stuff,” organizations focus on action and take a rather laissez-faire approach to project management, with little documentation and only loosely defined deliverables, timelines and budgets. On the other end

A N A LY T I C S - M A G A Z I N E . O R G

of the spectrum, labeled “Buttoned Up,” organizations take a disciplined approach to planning, monitoring and executing projects, with liberal documentation and carefully defined deliverables, schedules and budgets. This approach is common among external consulting groups by necessity, since clients insist on having answers to the three key questions of project management: what is to be delivered, W W W. I N F O R M S . O R G

when and at what cost? The former approach is more common among internal consulting groups, especially those with corporate funding, which provide internal clients with what appear to be “free” resources. Why should analytics professionals care about project management? After all, we are trained problem-solvers and when we “do stuff,” good things happen, right? On the contrary, projects are the primary vehicles through which we deliver value, and consistently delivering value with analytics requires effective project management. When When it comes to delivering analytics, what matters is not only what you do, but also how it comes to delivering analyt- you do it. Photo Courtesy of 123rf.com | alphaspirit ics, what matters is not only what you do, but also how you do it. “intended” is significant; projects fail to A project can be defined as “a piece achieve desired aims more frequentof planned work or activity that is comly than we would like to admit. Project pleted over a period of time and intended management provides an important mitito achieve a particular aim” [1]. Analytics gation strategy to improve the odds for projects begin with a purpose or aim – for success. What is your level of familiarity example to describe, to predict or to prescribe. Achieving that purpose requires activity or work, work that takes time and consumes scarce resources. The word Figure 1: Approach to project management spectrum. A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

PROJ E CT MA N AG E ME N T with the following basic project management concepts? 1. Project phases 2. Work breakdown structure (WBS) 3. Issue management 4. Risk management 5. Communication plan VALUABLE RESOURCES In case your project management IQ could use a boost, take advantage of the valuable resources – reference materials, training, peer groups and conferences – available through the Project Management Institute (PMI), the world’s premiere project management organization [2]. In addition, befriend experienced project managers (PMs) who can provide valuable advice. One PM taught me to use discrete percentages when reporting progress against project activities: 0 percent if not started, 20 percent when started, 80 percent when complete except for “loose” ends, and 100 percent when completely finished. Another suggested that when planning a project, focus less on task duration and more on when tasks can be completed, treating completion dates as contractual and allowing individual team members to manage their own schedules and responsibilities. Analytics projects are just that, projects, so general project management principles apply. And yet, analytics

A N A LY T I C S - M A G A Z I N E . O R G

projects include unique aspects that require tailored approaches. For one, analytics projects involve specialized subject matter. Choosing an appropriate technique, defining corresponding tasks, estimating level of effort for those tasks and executing the project require analytics expertise. For another, analytics projects include particular types of uncertainty. Data availability and quality is frequently an issue. Knowing what will happen when specific data meets math is unpredictable. Textbook techniques seldom apply directly to real-world problems, giving rise to the need for invention within analytics projects. Finally, change management can be especially challenging given the general level of discomfort among people with mathematics and math-based solutions. A project management methodology provides a generic set of process steps that can be customized for a specific project. Leveraging a methodology eliminates the need to reinvent the wheel for each new project. Two methodologies are especially relevant for analytics projects – CRISP-DM and Scrum. While CRISP-DM or Cross Industry Standard Process-Data Mining was developed by an industry consortium to support data mining projects, the methodology readily translates to any analytics project. Scrum,

W W W. I N F O R M S . O R G

a form of agile, was originally developed to support software development projects, but it also applies well to analytics projects. As shown in Figure 2, CRISPDM includes six phases â&#x20AC;&#x201C; Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation and Deployment [3]. While these phases may be executed one after another in a linear manner, the intent of the methodology is to support iteration. For example, after evaluating a model, a modeler might conclude that additional data is needed, giving rise to the need for additional business Figure 2: CRISP-DM phases. understanding, data understanding and data preparation. Deployment people. While the analytics â&#x20AC;&#x153;swim laneâ&#x20AC;? can involve implementing a model and/or was planned using CRISP-DM, the overpublishing findings and recommendations all project was planned using a software depending on the nature of the project. development methodology. The analytics The outer ring of the figure represents the swim lane was synchronized with other cyclical nature of the lifecycle of analytics pertinent swim lanes (primarily data and models [3]. system) according to what was required and when with respect to required inputs DATA MINING PROJECT and model outputs. The author applied CRISP-DM in a CRISP-DM worked well as an apdata mining project to develop anomaly proach within the analytics team. With its detection models for mining machine seniterative nature, though, the methodolsor data (i.e., data mining on mining data). ogy posed challenges to track and report With a staff of about 15 people, analytics progress within the overall project. For was a relatively small part of the overall example, the project manager found it project, which included more than 100 confusing to learn that the analytics team

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

PROJ E CT MA N AG E ME N T had looped back to data understanding. In retrospect, conducting an initial round of business understanding and data understanding would have made sense, followed by time-boxed modeling iterations which encompass business understanding, data understanding, data prep, modeling and evaluation as needed. Scrum is a form of agile that addresses complexity and risk by creating a product (e.g., an analytics model) incrementally through a series of short iterations known as sprints and teasing out requirements based on feedback from concrete deliverables [4]. Requirements are captured as stories of the form “a user with a certain role wants to use the system to …” in a list known as the project backlog. At the beginning of each sprint, the backlog is prepped and a sprint planning meeting conducted to select which stories will be pursued during the sprint based on priority, required level of effort and available development resources. During the sprint, the development team conducts a daily Scrum to review work completed the previous day and to be completed during the current day. At the tail

Figure 3: Scrum sprint cadence.

A N A LY T I C S - M A G A Z I N E . O R G

end of each sprint, a tested, “shippable” version of the product is demonstrated during a sprint review meeting. The project backlog is updated based on feedback captured during the review meeting to initiate the next sprint (see Figure 3). ANALYTICS PROJECT The author applied Scrum to an analytics project to develop integrated simulation and optimization models for planning oil pipeline terminal infrastructure. A team of four executed four three-week sprints. Despite lack of experience with Scrum, the team caught on quickly. The value of Scrum emerged during the first two sprint review meetings. The team expected and received feedback on its partially complete models. Unexpectedly, the team witnessed sidebar conversations among client employees, an airing of disagreements that arose in response to the demonstration of tangible models. Resolving these disagreements contributed both to better models and to customer buy-in. The team found that three weeks was long enough to deliver significant new capabilities but short enough to provide little room for delays. By reviewing required weekly progress at the beginning of each week, the team was able to stay on track during each sprint. Having ready

W W W. I N F O R M S . O R G

access to subject matter experts was critical to avoid delays. The team adapted Scrum to its needs, for example, by using only one simple metric (stories completed) to track progress. Being flexible with project management is critical to success. Adjust the approach based on the size and complexity of a project and the trade-off between the cost and benefit of various practices. MONITORING AND CONTROLLING In the realm of project management, project plans receive an outsize amount

of attention. Planning a project is certainly important, but monitoring and controlling during execution is equally important for success. At sufficiently frequent (e.g., weekly) review intervals, the team reports progress as well as issues. Successful organizations create an environment in which issues can be raised without risk of judgment or sanction. The sooner an issue is surfaced, the sooner that issue can be addressed, often by tapping into higher-level powers through escalation. Adherence to schedule, scope and budget are tracked carefully during

THE ODDS ARE IN YOUR FAVOR AT ANALYTICS 2017! Exhibitor and Sponsorship Opportunities at the 2017 INFORMS Conference on Business Analytics and Operations Research. This premier Analytics conference draws 800+ practitioners for three days of networking, professional development & intensive learning. For Sponsorship and Exhibitor details visit: http://meetings.informs.org/analytics2017

Contact: Olivia Schmitz INFORMS Exhibit & Sponsorship Sales Manager Olivia.Schmitz@informs.org 443.757.3539

CAESARS PALACE, LAS VEGAS APRIL 2-4, 2017

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

PROJ E CT MA N AG E ME N T project execution, and non-minor adjustments are made through a formal project change process. Accurately estimating the number of resources and length of time to accomplish tasks is critical to project success. Underestimating required level of effort may result in project delays or costly overruns. Overestimating may result in a plan being rejected by the client since the project takes too long or costs too much. When it comes to estimation, experience matters, so shadowing a seasoned analytics professional is helpful. So, too, is iteratively breaking down tasks into more easily estimated tasks. For example, to estimate how long “business understanding” will take, ask how this phase will be accomplished. Suppose that one of the business understanding approaches is to conduct a workshop. Now, conducting a workshop involves preparation and follow-up in addition to the workshop itself. Preparation might require two days, the workshop one day and follow-up two days for a total of five. Accurately estimating the time to conduct a workshop is relatively easy, while estimating the time to conduct the high-level, intangible activity of Business Understanding is not. In some cases, conducting a small pilot to understand the level of effort required for a task makes sense. For example, if custom analytic

A N A LY T I C S - M A G A Z I N E . O R G

models are to be created for each of 100 manufacturing plants globally, then the model might first be created for one or two (perhaps the most complex) in order to understand overall effort needed. The INFORMS Analytics Maturity Model assesses an organization’s analytics maturity in terms of three groups of four factors in the areas of organization, analytics capability and data & infrastructure [5]. While project management is not specifically referenced in that model, a disciplined approach to project management is most certainly necessary for organizations to achieve analytics maturity, enabling the consistent delivery of quality work products. Adopting a more buttoned-up approach to project management rather than simply doing stuff can help your organization mature and achieve maximum return on its analytics investment. ❙ Erick Wikum (ewha755@yahoo.com) is treasurer of the Analytics Society of INFORMS. For nearly 25 years, he has planned, conducted, monitored and led analytics projects to improve decision-making as both an internal and external consultant.

REFERENCES 1. http://dictionary.cambridge.org/us/dictionary/ english/project/ 2. http://www.pmi.org/ 3. https://www.the-modeling-agency.com/crisp-dm.pdf 4. http://scrummethodology.com/ 5. https://www.informs.org/Apply-OperationsResearch-and-Analytics/Analytics-Maturity-Model/

W W W. I N F O R M S . O R G

CROW D SOU RC I NG

Using the crowd: curated vs. unknown

BY BEN CHRISTENSEN

ou’ve heard from colleagues, from industry news, maybe even from personal experience of the success of crowdsourcing. With nearly half the world’s population online, it makes sense to tap into this tremendous resource to collect large quantities of data by breaking the collection down into micro-tasks that the enormous crowd of Internet users will complete for you at pennies per task. But you’ve probably also heard of cases of crowdsourcing gone wrong, cases like the British

A N A LY T I C S - M A G A Z I N E . O R G

government agency that put the task of naming its new research vessel out to the crowd and ended up with Boaty McBoatface [1]. So you have a human annotation task – a data refinement or evaluation task that requires human input – and you’re trying to decide whether to take your chances with crowdsourcing. Where do you start? The first thing you need to know is that you have options; it’s not Boaty McBoatface or nothing. “When should I use crowdsourcing, and when should I use a curated crowd?” This is the question anyone interested in

W W W. I N F O R M S . O R G

The power of crowdsourcing is in its numbers. Many hands make light work. Photo Courtesy of 123rf.com | Kheng Ho Toh

staffing human annotation tasks should be asking, but many don’t because they don’t even know there are two different options. So let’s start there – defining the options. Assuming you need human annotation, for example for search relevance evaluation, there are two ways you can gather the necessary humans to do that work: 1) crowdsourcing, where the task is made available to a large crowd without any management beyond a very limited set of task instructions and possibly a simple screening test; or 2) curated crowds, where a smaller group is selected to complete the task accurately according to quality guidelines.

The power of crowdsourcing is in its numbers. You can accomplish a lot quickly

because many hands make light work. A hundred thousand people can do quite a bit more than a hundred can. The cost is less because crowdsourcing typically pays only a few pennies per task. Most members of the crowd aren’t trying to make a living – they’re just trying to make a few extra bucks in their spare time. There’s usually little overhead involved in crowdsourcing because the crowd looks after itself. You put the task out there, and if it’s interesting enough and pays enough, the crowd will get it done. This model works well for simple tasks that require little explanation and even less expertise. For example, you can ask the crowd to choose which of two images contains a dog, to tell you whether a business listing is or isn’t a restaurant, or to transcribe words from

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

WHEN TO USE A TRADITIONAL CROWDSOURCING MODEL

CROW D SOU RC I NG

images [2] and have a reasonable amount of success. Opinion-based tasks where you are looking for a wide variety of responses are also well-suited for crowdsourcing: which image do you like better, what’s the best Italian restaurant in your town, or how would you word this request for a voice-activated personal assistant. CROWDSOURCING CHALLENGES With the advantages of traditional crowdsourcing come a few limitations. First, quality control is minimal. Without this, you must rely on clear instructions, automated understanding checks and high overlap to get data you can trust. Overlap is important because there will always be noise in the crowd – bad data that you have to identify and sift out – so you’ll likely pay for at least five members of the crowd to review each result. Some members of the crowd will try to game the system, using bots to do their work for them, so you’ll need to account for this with screening tests on top of your high overlap. The second limitation is your lack of control over task completion. The crowd can get a lot done quickly, but it will only get your task done quickly if it wants to. This means if your task is more difficult or less exciting than the shiny new task another team is offering, 44

A N A LY T I C S - M A G A Z I N E . O R G

then you’re going to have to incentivize the crowd with higher payment to work on your task. The crowd has made no commitment to you and may not be motivated to make a deadline. Finally, if you’re looking for data in smaller markets, you may be out of luck – crowdsourcing is huge in the United States and a few other countries, but the same doesn’t hold true globally. THE ALTERNATIVE: USING A CURATED CROWD Curated crowds, on the other hand, are all about quality. With this solution, offered by a small number of specialized crowd data solution providers, you have a group of people who are dedicated, if not specifically to your task, then to similar tasks. People in curated crowds become experts in search relevance evaluation, social media evaluation or whatever type of human annotation tasks they work on. This is not simply a case of counting on their accumulated experience to ensure quality, although that experience does play a large role. The key to quality is constant checks and balances. They are held to quality metrics, receive quality feedback, and are removed from your task if they don’t deliver the required quality. This means that you can use very little overlap, paying for each judgment W W W. I N F O R M S . O R G

only one to three times instead of five or more times, because you can trust the data each person delivers. Curated crowd providers also monitor productivity and throughput, ensuring that the crowd meets their weekly, daily and hourly commitments so that you have the data you need when you need it. And if you’ve chosen a good vendor, then the manager will also be an invaluable resource, leveraging years of experience to partner with you in building out tasks and guidelines based on your needs. With this higher level of quality and productivity management comes a cost. Curated crowds cost more than crowdsourcing because this work is typically a primary source of income. You also pay for the quality oversight that you don’t have in crowdsourcing. Keep in mind, though, that lower overlap mitigates these costs because you aren’t paying for each collected data point multiple times. Apart from the financial cost, curated crowds also require more of a commitment from you in exchange for the greater commitment you get. The curated crowd will be happiest and will keep their skills sharpest when you provide work for them consistently. That said, if the natural ebb and flow of your need for human annotation necessitates more flexibility, there are

alternatives, such as sharing a flexible curated crowd with other teams running similar tasks.

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

USE CASES: SEARCH AND SOCIAL Major worldwide search engine providers have been using curated crowd solutions for years. Curated crowds are used for search relevance evaluation, local search result validation, query classification, spam identification and countless other tasks that require more attention than what traditional crowdsourcing provides. By using this model,

CROW D SOU RC I NG

these search engine providers gather high-quality data they can trust to accurately measure the success of their current algorithms, compare their search engine against competitors, and test out new iterations before launching. Social media network providers have more recently come to appreciate the value of the curated crowd. While it’s common for these providers to poll their users regarding their experience with the site, the social feed, the ads and the search functionality, the data gathered from these traditional crowd-based methods is limited and uncontrolled. By contrast, engaging a curated crowd that is able to provide targeted feedback on specific aspects of the social feed, filtering their subjective experience through a set of objective criteria, has produced much more useful data. Social media providers who take advantage of this model are able to leverage the resulting data to improve their social feed algorithms, their ads and their search features in order to create a user experience that stands out above their competitors. CHOOSING THE RIGHT OPTION So when should you use crowdsourcing and when should you use a curated crowd? Crowdsourcing is great for simple tasks that can be adequately 46

A N A LY T I C S - M A G A Z I N E . O R G

explained in two or three sentences. You’ll get a lot done quickly, but be prepared to raise the pay rate if you have a tight deadline and the crowd doesn’t find your task sexy enough. On the other hand, if you have a more complex task, particularly if it’s a longer-term or ongoing task that dedicated people can build expertise on over time, then a curated crowd is for you. Either way, be sure you fully understand your options so that you can make the best choice for your business. And if you’re trying to name a boat, you might want to limit the vote to names you won’t be embarrassed to paint on the side of that brand new vessel. ❙ Ben Christensen, director of content relevance operations at Appen, has been managing crowdbased search, eCommerce and social evaluation work since 2008. He has a master’s degree in library and information science from the University of Washington. Appen is a global language technology solutions provider with capability in more than 180 languages and 130 countries, serving companies, automakers and government agencies.

REFERENCES 1. http://www.nytimes.com/2016/03/22/world/europe/ boaty-mcboatface-what-you-get-when-you-let-theinternet-decide.html?_r=0 2. This is exactly what the Gutenberg Project does: https://digitalciv.wordpress.com/2010/10/26/ crowdsourcing-captchas-and-the-gutenbergproject/

W W W. I N F O R M S . O R G

CROP CHALLENGE in ANALYTICS

The 2017 Syngenta Crop Challenge in Analytics focuses on the seed retailer, who sells soybean seed varieties to farmers. The farmers require different soybean seed varieties based on expected growing conditions. To maximize yield in a region with a large number of farmers, the retailers need to predict and stock the soybean variety seeds that will thrive best in the farmer’s most common growing conditions. It is difficult for a seed retailer to predict which seed varieties to stock almost a year in advance to the soybean crop planting by the farmers.

THE CHALLENGE:

Which soybean seed variety, or mix of up to five varieties in appropriate proportions, will best meet the demands of farmers in a growing region?

CHALLENGE LAUNCH - Data Available NOW! • Deadline for submissions January 16, 2017 • Selection of finalists February 24, 2017 • Finalist presentation (live or via telepresence) at INFORMS Conference on Business Analytics and O.R. April 2–4, 2017

SO FTWARE S U R VE Y

Decision analysis Past, present and future of dynamic software emphasizes continuous improvement of vital analytics tool.

BY SAMANTHA OLESON

ifty years have elapsed since the founding of the field of decision analysis by Howard Raiffa [1] and Ron Howard [2]. 2016 is not only a milestone year due to the anniversary but also because it marks the passing of one of the founders, Howard Raiffa. Such significant events make this year a time for reflection. Seasoned decision analysts can celebrate their contributions to the advancement of the field, while young operations research (O.R.) and analytics professionals should take this time to imagine how their current and future contributions will shape the next 50 years of the field. 48

A N A LY T I C S - M A G A Z I N E . O R G

In a 1988 paper on the state of the field of decision analysis, Ron Howard wrote, “the accomplishments and promise of the field are impressive” and improvements to the “procedures for formulating, eliciting, evaluating, and appraising the decision problem” occur every day. He further explained that despite the improvements, decision analysis has “not [yet] become commonplace even in very important decisions.” However, he believed that as of 1988, “decision analysis [was] poised for a breakthrough in its usefulness to human beings” that would be achieved in part using “[computer-based] intelligent decision systems … to provide the benefits W W W. I N F O R M S . O R G

of decision analysis on a broader scale than ever before” [3]. Now, 30 years after the 1988 article, we can agree with Ron Howard that decision analysis has accomplished much since its founding. He was also correct in predicting that technology would play an ever-increasing role in the decision analysis field and that more contributions are still to come. Although a very difficult task, it is possible to identify several of these contributions as defin- How to make the visions of the founders of decision analysis and current leaders a reality? Photo Courtesy of 123rf.com | Sergey Nivens ing achievements for the first 50 years of the field of decision analysis. 1988 predictions in that they identified These include: two factors that would characterize • establishment of university programs the future of decision analysis: 1) and classes focused on decision collaboration with other fields of study analysis, and, 2) advances in technology [4]. Eric • adoption of decision analysis Horvitz, managing director of Microsoft techniques and principles in Research’s Redmond lab, cited the everyday business practices, and role of “decision-theoretic ideas” in the • integration of computing methods development of artificial intelligence and tools. as a prime example of the significant contributions the field of decision analysis During a 2011 meeting of members has made in the past and will make in of the Decision Analysis Society of the future. He went on to conclude that, INFORMS, leading decision analysts “There is great opportunity for more contemplated the future of the field. Their interaction between the communities conclusions mirrored Ron Howard’s … to date the surface has only been A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

DE CIS IO N A NA LYS I S S U R VE Y

scratched” [5]. Many attendees agreed that future work in decision analysis will not only prove the practice to be indispensable to the field of computer science but also to “such realms as healthcare, climate change, energy and national security” [4]. Now, standing on the threshold of the next 50 years of decision analysis we must ask ourselves, “How can we make the visions of our founders and current leaders a reality?” Helping to tackle the present and future challenges of interfield collaboration and integration of advanced computing methods are the vendors that develop the decision analysis tools used by the O.R. community today. This year’s software survey seeks to catalog these tools and the features they offer to practitioners. The goal is to help readers to use these tools to continue to further spread the use of the field of decision analysis. THE SURVEY This 2016 software survey assists readers in evaluating the featured decision analysis software products in three categories: 1) decision analysis applications, 2) usability features, and 3) licensing and training options. Decision analysis applications examine the analytical uses of the tools, as well as 50

A N A LY T I C S - M A G A Z I N E . O R G

features offered for elicitation of decision problem components. Usability features highlight available features that make the tools user-friendly, compatible with other software and operating systems, and useful in communicating results. Since the 2014 survey, we added questions to this category to more closely examine this area. The final category, licensing and training options, asks for the options provided by the vendors for purchasing and achieving proficiency with the products. The approach and collection method of this year’s survey did not differ from previous years. Vendor representatives completed an online questionnaire consisting of approximately 60 questions. Vendors who completed the 2016 survey include those who participated in previous surveys or those who recently came to the attention of Lionheart Publishing staff. Vendors who have not yet participated in the 2016 decision analysis software survey may submit details of their product to the online version of the survey by filling out the questionnaire available at http://lionhrtpub.com/ ancill/dassurvey.shtml. Results of the 2016 software survey (URL) are presented verbatim. The results do not necessarily imply quality or cost effectiveness, but rather to provide a detailed catalogue of possible W W W. I N F O R M S . O R G

decision analysis tools available on the market today. 2016 RESULTS The 2016 software survey features 29 software packages from 19 vendors. Companies from the United States, United Kingdom, Belgium, Finland, Canada, New Zealand and Sweden are represented in this yearâ&#x20AC;&#x2122;s survey. These companies provide their decision analysis packages to a variety of industries, with the healthcare, defense, energy and mining industries reported most frequently.

While not all of the vendors who participated in 2014 responded this year, first-time responders submitted information for eight software packages. Three of these new responses were for products just launched in 2016. The following sections provide a brief overview of the results of each of the three survey categories: Decision analysis applications: Vendors most frequently reported that their products could be applied to decision problems involving multiple competing objectives and uncertainty. A smaller

SAVE THE DATE

HEALTHCARE 20 7

OPTIMIZING OPERATIONS & OUTCOMES July 26â&#x20AC;&#x201C;28, 2017 Rotterdam, Netherlands

HTTP://meetings.informs.org/healthcare2017

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

DE CIS IO N A NA LYS I S S U R VE Y

number of products offered features for risk tolerance, sequential decisionmaking and evidential reasoning. Other features reported by vendors that were not specifically queried in the survey included Monte Carlo simulation, tradeoff analysis, group value elicitation, game theory and decision framing. Vendors reported that more than 75 percent of tools could be used for elicitation of value functions/scores and model structure. Decision analysis features less frequently offered by the products included elicitation of probabilities, criteria/attribute weighting and the value of imperfect information. Usability features: Not surprisingly, most of the products provide their users with common usability features such as the ability to import and export components, document model structure and/ or judgments with text and display analytical results graphically. Additionally, vendor responses show that nearly 90 percent of the decision analysis tools featured are capable of interfacing with other software, and approximately 50 percent have XML and/or API features. Thirteen of the 29 packages are offered as web implementations. Of the 13, three are veteran products providing web-based access as a new feature and two are software just released in 2016. 52

A N A LY T I C S - M A G A Z I N E . O R G

Licensing and training: Most of the responding vendors reported that their products can be purchased for either educational or commercial use. Nearly 35 percent of the products offer an enhanced or high-performance version of the tool. Many of the products are offered free or at a discounted rate for educational use, while the price of enhanced or high-performance and commercial versions of the tools vary depending on the licensing plan. All but five vendors reported that training programs are available to users either through the vendor itself or through a third party. Fifteen of these vendors offer web-based training, two of which are returning products that did not previously offer web-based training. CURRENT AND FUTURE TRENDS Although it is difficult to conduct detailed statistical analysis of the data set due to the small sample size and changes in the list of participants year to year, it is possible to identify several trends: Collaboration: With each year of the survey, vendors report the expanded capability of their tools to provide collaborative decision analysis solutions. Todayâ&#x20AC;&#x2122;s tools not only feature group collaboration capabilities such as group elicitation, and simultaneous data input and viewing, but also features that lay the groundwork for inter-field collaboration. W W W. I N F O R M S . O R G

Compatibility with other software and/ or XML and API features make decision analysis techniques more readily accessible to other industries and fields of study. The forward-thinking development and continuous improvement of decision analysis tools will help users to realize the collaborative future envisioned by both Ron Howard in 1988 and by the 2011 workshop attendees. Web implementation and cloud computing: Continuing the trend from 2014, a growing number of vendors offer web-based implementations of their

products and web-based training and support. In fact, two of the respondents this year specifically stated that their packages are cloud-based products. With the increasing use of “the cloud,” it is likely that products will be offered in this environment more and more frequently. Visualization: Improvements in computing technology have made it easier than ever to interact with the decision space through the data and values that define it. With this in mind, this year’s survey introduced new questions that examine the communication and visualization

! PYNE O C LI

URable ONrs! O Y ail be

v m S ES Rials aMS Me C C tO R

u 6 T INFO 201 for

Optimization Challenges in Complex, Networked, and Risky Systems 2016 Optimization Challenges in Complex, Networked, and Risky Systems Aparna Gupta and Agostino Capponi, Tutorials Co-Chairs and Volume Editors J. Cole Smith, Series Editor The TutORials in Operations Research series is published annually by INFORMS as an introduction to emerging and classical subfields of operations research and management science. These chapters are designed to be accessible for all constituents of the INFORMS community, including current students, practitioners, faculty, and researchers. The publication allows readers to keep pace with new developments in the field, and serves as augmenting material for a selection of the tutorial presentations offered at the INFORMS Annual Meeting.

INFORMS 2016 edition of the TutORials in Operations Research series will be available online to registrants of the 2016 INFORMS Annual Meeting on November 12, 2016. The fifteen chapters in this year’s volume highlight the tutorial theme of Optimization Challenges in Complex, Networked, and Risky Systems. The chapters are written by a diverse array of experts working across a variety of institutes. Their research covers a range of exciting topics, including multiobjective optimization, optimization under uncertainty, big data analytics, project management, and risk modeling and optimization. The chapters focus on many compelling applications for which this analysis is useful, including those arising in finance, healthcare, and energy systems.

Access the 2016 TutORials at:

http://pubsonline.informs.org/series/educ

Aparna Gupta and Agostino Capponi, Tutorials Co-Chairs and Volume Editors

J. Cole Smith, Series Editor

Optimization Challenges in Complex, Networked, and Risky Systems

2016

Aparna Gupta and Agostino Capponi, TutORials Co-Chairs and Volume Editors

www.informs.org

Nashville 2016 Presented at the INFORMS Annual Meeting, November 13– 16, 2016

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

DE CIS IO N A NA LYS I S S U R VE Y

features offered by the various products. Vendors reported that about 75 percent of tools have customizable visualization features that allow users to perform actions such as drag-and-drop chart development and manipulation, or color and formatting selection. As for decision space exploration using interactive graphics, displays, or interfaces, about 70 percent of the products surveyed had this capability. While visualization and communication may not be the primary focus of tools used for decision analysis, the number of vendors reporting significant updates to the reporting, graphing and formatting features of their tools suggest they understand that there will be a strong need for these features in years to come. FINAL THOUGHTS In 2016, we not only celebrate the accomplishments of Ron Howard and Howard Raiffa but also the accomplishments of the many decision analysts who used their founding concepts to expand the field even further. As a young decision analyst, it is difficult to imagine that the founders left much room for further innovation. How are we supposed to continue to advance the field over the next 50 years? Leading voices in decision analysis say there is room to expand through 54

A N A LY T I C S - M A G A Z I N E . O R G

collaboration with other fields and through the application of advanced computing methods, so we must focus our energies in these areas. There is much uncertainty associated with the roles we will play in the future of decision analysis, however, this is the type of problem we have been well trained to overcome. It’s time to leap into the next 50 years. ❙ Samantha N. Oleson (soleson@innovativedecisions. com) is an analyst with Innovative Decisions, Inc., an analytics consulting firm specializing in decision and risk analysis, operations research and systems engineering.

EDITOR’S NOTE: A version of this article appeared in OR/MS Today, the membership magazine of the Institute for Operations Research and the Management Sciences (INFORMS).

REFERENCES 1. Gavel, Doug, 2016, “Harvard Remembers Howard Raiffa,” Harvard Gazette, News.harvard.edu, July 11, 2016. Web access: Sept. 7, 2016. 2. Stanford University, 2016, “Profiles: Ronald Howard,” accessed Sept. 9, 2016. 3. Howard, Ronald A., 1988, “Decision analysis: practice and promise,” Management Science, Vol. 34, No. 6, pp. 679-695. 4. Abbas, Ali, 2012, “Decision analysis: past, present and future,” OR/MS Today, February 2012. 5. Horvitz, Eric, Abbas, Ali, 2012, “Decision Analysis Workshop.” Procedures of Decision Analysis Workshop, Palo Alto, Calif., OR/MS Today, February 2012 (print version).

SURVEY DATA & DIRECTORY To view the 2016 decision analysis software survey data, click here. To view the 2016 directory of decision analysis software vendors, click here.

W W W. I N F O R M S . O R G

Is the largest association for analytics in the center of your professional network? It should be. • INFORMS Allows You to Network With Your Professional Peers and Others Who Share Your Interests • INFORMS Connect, the New Member-only, Online Community Lets You Network With Your Colleagues • Unsurpassed Networking Opportunities are Available in INFORMS Communities and at Meetings • INFORMS Offers Certification for Analytics Professionals • Take Leadership Roles to Help Build Your Professional Profile • INFORMS Career Center Provides You With the Industry's Leading Job Board

Join Online Today! http://join.informs.org

DATA L AK ES

The biggest big data challenges Why data lakes are an important piece of the overall big data strategy.

BY PRASHANT TYAGI (left) AND HALUK DEMIRKAN today’s complex business world, many organizations have noticed that the data they own and how they use it can make them different than others to innovate, to compete better and to stay in business [1]. That’s why organizations try to collect and process as much data as possible, transform it into meaningful information with datadriven discoveries, and deliver it to the user in the right format for smarter decision-making [2]. Big data analytics has become a key element of the business decision process over the last decade.

A N A LY T I C S - M A G A Z I N E . O R G

With the right analytics, data can be turned into actionable intelligence that can be used to help make businesses maximize revenue, improve operations and mitigate risks. Big data is the term for a collection of data sets so large and complex that it becomes difficult to process using hands-on database management tools or traditional data processing applications. The challenges include capture, curation, storage, search, sharing, transfer, analysis, visualization and many other things. According to Demirkan and Dal [2], big data has the following six “V” characteristics:

W W W. I N F O R M S . O R G

• Volume (data at rest): terabytes, exabytes, petabytes and zettabytes of data • Velocity (data in motion): capturing and processing streaming data in seconds or milliseconds to meet the need or demand. • Variety (data in many forms): structured, unstructured, text, multimedia, video, audio, sensor data, meter data, html, text, emails, etc. • Veracity (data in doubt): According to Gartner, 60 percent of companies say they don’t have the skills to make the best use of their data. uncertainty due to Photo Courtesy of 123rf.com | Bruce Rolff data inconsistency and incompleteness, ambiguities, latency, IDC predicts revenue from the sales deception, model approximations, of big data and analytics applications, accuracy, quality, truthfulness or tools and services will increase more trustworthiness. than 50 percent, from nearly $122 • Variability (data in change): the billion in 2015 to more than $187 billion differing ways in which the data in 2019 [3]. Even though 73 percent of may be interpreted; different companies intend to increase spending questions require different on analytics and making data discovery a interpretations. Data flows can be more significant part of their architecture, highly inconsistent with periodic 60 percent feel they don’t have the skills peaks. to make the best use of their data [4]. • Value (data for co-creation): The Given an abundance of knowledge and relative importance of different data experience, combined with successful to the decision-making process. data and analytics-enabled decision

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

DATA L AK ES

DATA-RELATED CHALLENGES FOR BIG DATA Slow on boarding and integrating data

support systems, big data initiatives come with high expectations, and many of them are doomed to fail. Research predicts that half of all big data projects will fail to deliver against their expectations [5]. When Gartner asked what the biggest big data challenges were, the responses suggest that while all the companies plan to move ahead with big data projects, they still don’t have a good idea as to what they’re doing and why [6]. The second major concern is not establishing data governance and management [7] (see Table 1). THE DATA LAKE JOURNEY In the context of governance and management of big data, the term “data lake” has been widely discussed in recent years. Is a data lake a logical data warehouse to manage the six Vs of big data? How do data lakes relate to data warehouses? How do data lakes help evolve the data management architecture and strategy in organizations? The data lake concept has been well received by enterprises to help capture and store raw data of many different types at scale and low cost to perform data management transformations, processing and analytics based on specific use cases. The first phase of the data lake growth was in consumer web-based companies. The data lake strategy has already shown positive

A N A LY T I C S - M A G A Z I N E . O R G

Transforming data from growing variety of technologies Custom coded ETL ETL processes are not usable Optimizing for analytics is costly and time consuming Wait for need of a defined data set, wait for data on boarding Poorly recorded data provenance Data meaning lost in translation Data transformations tracked in spreadsheets Post on boarding maintenance and analysis cost is high Recreating lineage is manual and time consuming Help consume difficult target data Optimization favors known analytics not suited for new requirements One size fits all canonical view rather than fit for purpose views Lacks a conceptual model to easily consume target data Difficult to identify what data is available, getting access and integration Industrializing the big data environment which is difficult to manage Data Silos lead to inconsistency and sync issues Conflicting objectives of opening access yet managing security and governance Rapid biz change invalidate data org and analytics optimizations Managing the integration / interaction with multiple DM tech in Big Data Environment

Table 1: The unique data-related challenges for big data.

results for these consumer businesses by helping increase speed and quality of web search, web advertising (click stream data) and improved customer interaction and behavior analysis (cross-channel analysis). This led to the next phase for data lakes, which was to augment enterprise data warehousing strategies.

W W W. I N F O R M S . O R G

A traditional data warehouse supports batch workloads and simultaneous use by thousands of concurrent users performing basic reporting and advanced analytics with a pre-defined data model. However, a lot of cleaning and other work is required before the data is properly captured and ready for modeling. A data lake, on the other hand, is meant to provide faster ingestion of raw data and to execute batch processing at scale on the data. The warehouse schema means data must be captured in the code for each program accessing the data. Given the capability of a data lake to ease the ingestion and transformation processes, it became a natural choice for migrating the extract, transformation and loading (ETL) workloads off traditional warehouses in order to provide a scale-out ETL for enterprises and big data. This makes a data lake suitable for data ingestion, transformation, federation, batch-processing and data discovery. The implementation characteristics of a data lake, namely inexpensive storage and schema flexibility, make it ideal for insight discovery. However, these traits do not necessarily translate to a high-performance, production-quality analytical platform. Making new insights available to the broadest possible audience requires data optimization, greater maturity of analytical models and semantic consistency. As new insights are discovered, the work passes

from the data science team to the data engineering team. Data engineers take the new questions and optimize for new answers. They refine and optimize the raw data, as well as the analytical models. Existing data integration processes can be used, or new processes can be built.

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

DATA LAKE IMPLEMENTATION EXAMPLES GE Predix is an industrial data lake platform that provides rigid data governance capabilities to build, deploy and manage industrial applications that connect to industrial assets, collect and analyze data, and deliver real-time insights for optimizing industrial infrastructure and operations. Following are sample use cases from two major industries: Aviation: Two factors in the aviation industry are key in determining the profitability for airline businesses: 1) the accuracy in prediction of fuel price fluctuations, and 2) predictive maintenance that can improve â&#x20AC;&#x153;time on wingsâ&#x20AC;? (the time an aircraft is actually in flight). The aviation group at GE analyzes data from more than several thousand engines per day to identify sub-optimal performance parts. This identification needs to be done almost instantly so it is a use case for real-time analytics. The data collected by sensors deployed on these engines capture physical parameters such as time, temperature,

DATA L AK ES pressure, etc. The correlation of this data and identification of insights through these correlations are accomplished by creating and using analytical models. An example of this is the analytics for large historical data using physics-based models such as numerical propulsion simulation system (NPSS) to check engine deterioration and other engine utilization algorithms. The use of a data lake enabled 10 times faster analytics capabilities, which meant that these algorithms and models could be executed in days (instead of months). The company learned that the hot and harsh environments in places such as the Middle East and China clogged engines, causing them to heat up and lose efficiency, thus driving the need for more maintenance. GE learned that if it washed the engines more frequently, they stayed much healthier. This information can save a customer an average of $7 million in jet airplane fuel annually because the engines are more efficient [8]. Electrical power: The electrical power sector has seen a significant increase in system complexity over the past several years. Data diversity and volume is on the rise, and so is the cost of traditional data management. This required a shift in data strategy to create a pervasive culture of data-driven insights at scale. For example, more than 20,000 users are on the big data platform for this vertical,

A N A LY T I C S - M A G A Z I N E . O R G

which generates or handles upward of several millions transactions a day. Power companies provide analytics on historical data extraction services upon customer request. These data requests typically are for a number of tags for a number of units over time and are executed at a fleet level. In order to serve the demand, the team needs to ingest all historical data from thousands of gas and steam turbines into a data lake from current operational systems. This is done by implementing a daily ingestion process to keep the big data platform and data lake current within 24 hours of operational data, which is used for generating the analytics. The use of a data lake has helped reduce clock time and touch time to extract data from weeks to hours. It also helped reduce the load on operational systems by freeing them of the analytics burden, thereby reducing the data kept in expensive SAN storage systems from over a decade to only a few months. This allows operational systems to run “in-memory” with improved performance. STAYING ON COURSE TO SUCCESS So what should your organization do to stay on course for a successful big data journey? Given the above discussion, the real questions become: Should traditional data warehouses continue to support operationalization and re-use of the data

W W W. I N F O R M S . O R G

DATA WAREHOUSES VS. DATA LAKES DATA WAREHOUSES

DATA LAKES

CONTROL

FLEXIBILITY

Structured, processed

Data

Schema-on-write

Processing

Structured/ semi-structured / unstructured, raw Schema-on-read

Well-defined schema

Schema

No schema (schema on read)

Expensive for large data volumes

Storage

Designed for low-cost storage

Less agile, fixed configuration

Agility

Highly agile, configure & reconfigure as needed

Well-defined usage

Usage

Future, experimental usage

Clean, trusted data

Raw data, frictionless ingestion

Mature

Control

Maturing

Mature

Security

Maturing

Mature

Governance

Maturing

Mature

Quality

Maturing

Mature

Human resource

Maturing

High

Cost

Low

Flexibility

High

Mature

Reliability

Maturing

Mature

Performance

Maturing

Low

Reusability

High

Single model of the truth

Integration and end user exercise

Upfront data preparation

Late data integration

Skills: Heavy IT reliance (Large IT teams: DBAs, Data Architects, ETL Developers, BI Developers, DQ Developers, Data Modelers, Data Stewards) & Less technical analysts

Skills: Self-service. More technical analysts, and IT manages the cluster and ingestion, but no IT involvement when working with data (data scientists, developer)

Features in Data Lake that add value • Common Data Model: Data is glued together by the business meaning rather than physical structures dictated by underlying technologies. Semantic models relate data by business meaning, is that being facilitated by data lake? • Flexible data model: No need to redesign database to extend or revise data model • Connecting external data: External data can be sourced with real-time values and delivered at query time • Federation and virtualization: Choice for which data to copy and which to leave in the source of truth. Models for all businesses supported by this data copy • Data Definition: Less change management as expertise is captured in data definition

Table 2: Data warehouses vs. data lakes.

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

DATA L AK ES

Data Ingestion Model driven

and provide more upstream business value capabilities like in-memory database, graph database, NO-SQL reports and other analytics capabilities? Or should the lake be expanded to become a data discovery platform and retire the data warehouses? The goal is to converge these platforms while adding more capabilities for handling all enterprise use cases exploiting IT and transactional data and thereby helping migration to cloud-based services. With that in mind, the data lake should not be designed as just a big data platform based on Hadoop; it should be designed using multiple technologies. These technologies will help offload and retire data warehouses by providing enterprise data warehouse-like capabilities to handle advanced workloads and data discovery. However, the strategy should not be to replace the data warehouses completely but only to move the analytics capabilities to the lake. The transactional/operational workloads for reporting and closing books, etc., should stay on traditional warehouses. That will help reduce the footprint on the more expensive warehouses and use them for what they are best suited for. Data governance is extremely important for success of big data projects. Based on the value of data, data lakes can be structured as: 1) governed data (e.g.,

A N A LY T I C S - M A G A Z I N E . O R G

Semantic tagging On-demand query Streaming Scheduled batch load Self service Data Management Data movement Data provenance Types (In memory, NoSQL, MR, columnar, graph, Semantic, HDFS) Data flow (governed data, lightly governed data, ungoverned data) Query Management Semantic search Data discovery Analytics directed to best query engine Capture and share analytics expertise Query data, metadata and provenance Data Lake Management Models (Biz unit data optimized to assist analytics) Data assets catalog (ontologies, taxonomies) Workflow (processes, schedules, provenance capture) Access management (AAA, group/role/rule/user-based authorization) Metadata

Table 3: Four components of data lakes.

key business data is being understood for ownership, definition, business rules and quality), 2) lightly governed data (e.g., data is being understood in regard to definition and lineage, but not necessarily controlled with respect to quality or usage), and ungoverned data (e.g., data is being only understood in regard to definition and

W W W. I N F O R M S . O R G

location. Ungoverned data may or may not physically exist in the data lake and may exist only in the data catalog as metadata pointers to external data) [9]. Based on the governance of data, the term “data reservoir” is now being used to describe the managed, transformed, filtered, secured, portable and potable data (fit for consumption of data). For every type of data governance, metadata is critically important for the success of big data projects. Organizations need to look at strategically investing in data lake architectures and implementations. A coordinated effort around a data lake can help bring the data strategy around big data and analytics together. Leveraging the data lake for rapid ingestion of raw data that covers all the six Vs and enable all the technologies on the lake that will help with data discovery and batch analytics. In order to complement the capabilities of data lakes, an investment needs to be made for data extracted from the lake, as well as in platforms that provide real-time and MPP capabilities. It is this entire eco system that needs to be put in place and executed to work in synergy, which will lead to all the promised benefits of the big data ecosystem. The data lake is a key piece of the overall data strategy, and not the one size solution for all data needs. Data lakes need to have four primary

components – data ingestion, data management, query management and data lake management (Table 3). ❙

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

Prashant Tyagi (prashant.tyagi@ge.com) is a director of IoT and analytics platforms at GE Software. Previously, he led the strategy and execution for Cisco’s Smart Services. He has a MBA from the Indian Institute of Management in Bangalore and a master’s degree in computer science from Clemson University. Haluk Demirkan (haluk@uw.edu) is a professor of Service Innovation and Business Analytics at the Milgard School of Business, University of WashingtonTacoma, and a co-founder and board director of the International Society of Service Innovation Professionals. He has a Ph.D. in information systems and operations management from the University of Florida. He is a longtime member of INFORMS. REFERENCES 1. Demirkan, H. and Delen, D., 2013, “Leveraging the Capabilities of Service-Oriented Decision Support Systems: Putting Analytics and Big Data in Cloud,” Decision Support Systems and Electronic Commerce, Vol. 55, No. 1, pp. 412-421. 2. Demirkan, H. and Dal, B., 2014, “Why Do So Many Analytics Projects Still Fail? Key considerations for deep analytics on big data, learning and insights,” Analytics magazine, pp. 44-52, JulyAugust 2014; http://analytics-magazine.org/ the-data-economy-why-do-so-many-analyticsprojects-fail/ 3. http://www.informationweek.com/big-data/bigdata-analytics/big-data-analytics-sales-will-reach$187-billion-by-2019/d/d-id/1325631 4. http://data-informed.com/gartner-researcherspredictive-analytics-to-gain-traction-in-business/ 5. http://www.forbes.com/sites/ bernardmarr/2015/03/17/where-big-data-projectsfail/#50ee6465264e 6. http://readwrite.com/2013/09/18/gartner-on-bigdata-everyones-doing-it-no-one-knows-why 7. http://www.ibmbigdatahub.com/blog/10-mistakesenterprises-make-big-data-projects 8. Winig, L., 2016, “GE’s Bıg Bet On Data and Analytıcs,” MIT Sloan Management Review, Feb. 18; http://sloanreview.mit.edu/case-study/ge-bigbet-on-data-and-analytics/ 9. https://infocus.emc.com/william_schmarzo/datalake-data-reservoir-data-dumpblah-blah-blah/

ANALY TIC S I N AC T I O N

The sport business analytics process

BY C. KEITH HARRISON (LEFT) AND SCOTT BUKSTEIN EDITOR’S NOTE: The following is an excerpt from the book, “Sport Business Analytics: Using Data to Increase Revenue and Improve Operational Efficiency,” co-edited by C. Keith Harrison and Scott Bukstein and recently published by Taylor & Francis (CRC Press). The book is part of a new book series on Data Analytics Applications edited by INFORMS member Jay Liebowitz.

he sport business analytics process generally involves data collection, management, visualization, implementation and evaluation. Sport business organizations are encouraged to focus first on clearly defining business strategies, goals and objectives before developing a data-driven initiative or staffing an analytics department. Next, organizations need to identify the data systems that will be used to collect and capture data. For example, a sport team could leverage ticketing and point-of-sale software

A N A LY T I C S - M A G A Z I N E . O R G

systems to monitor season ticket holder accounts (e.g., frequency of ticket utilization and most recent game attendance) and concessions sales (e.g., track food and beverage inventory along with corresponding revenue at each sales area). In addition to determining the “right” system(s) for data collection, it is imperative for organizations to access and assess the “right” data based on business strategy. For example, if a sport team plans to utilize intercept surveys to determine the probability of season ticket holder renewals, the team could focus on collecting the following

W W W. I N F O R M S . O R G

information from current season ticket holders: 1) amount spent on season ticket(s) and personal seat license (if applicable); 2) years of season ticket membership; 3) number of games attended during the current season; 4) whether season ticket holder is an individual or business; Analytics applications range from ticket pricing and sales inventory to customer relationship 5) distance of season ticket management and fan engagement. Photo Courtesy of 123rf.com | Oleksii Sidorov holder commute to each home game; 6) number of times season through the team’s official season ticket ticket holder attempted to resell tickets; 7) exchange program, and have not personsuccess rate with respect to season ticket ally attended a game in more than two holder attempts to resell tickets; and 8) atmonths. The team would likely flag these tendance/engagement at ancillary team season ticket holder accounts as “most events with exclusive access for season likely not to renew,” which could directly ticket holders. impact the renewal prioritization strategy An effective and efficient data manof team sales and service representatives agement system (i.e., “data warehouse”) responsible for renewing season ticket will enable a company to organize, stanaccounts. dardize, centralize, integrate, interconnect Data presentation and visualization will and streamline the collected data. An orthen empower analytics team members ganization will then be able to quickly mine to communicate results so that data is acthe data and create an analytic model that cessible, understandable and usable with transforms the raw data into practical, acrespect to developing operational stratetionable insight. For example, a sport team gies. After an organization implements could use Microsoft Excel or statistical softthe data-driven recommendations, key ware such as SAS to pinpoint all first-year decision-makers should consistently monseason ticket holders who purchased the itor and evaluate initiative effectiveness least expensive season ticket package, so that the organization can adjust both live more than 40 miles from the arena, business operations and future analytics have resold over 50 percent of their tickets processes. Sport business organization

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

S PO RT S AN A LY T I C S leaders should also continuously track industry best practices with respect to databased opportunities for collaboration and innovation. BUSINESS ANALYTICS APPLICATION AREAS Ticket pricing and sales inventory: Sport business organizations utilize analytics to inform the ticket inventory and pricing decision-making process. Most sport teams focus on a combination of “attendance maximization” and “revenue optimization.” Teams also focus on creating customer value (e.g., fan event experience) in addition to understanding the importance of “customer lifetime value” (e.g., cumulative amount of total business derived from a current or prospective ticket holder). Ticket demand models combined with direct feedback from customers assist sport organizations in developing ticket pricing strategies and customized ticket promotions. Customer relationship management and fan engagement: Sport organizations develop customer relationship management (CRM) systems both to create fan profiles and to structure ticket sales strategies. The CRM data warehouse functions as a centralized, integrated database for information related to customer demographics in addition to customer ticketing, merchandise, and food and beverage purchase patterns [1]. Organizations can then

A N A LY T I C S - M A G A Z I N E . O R G

analyze this data to develop customized messages for specific season ticket holders (or other categories of customers). For example, a college athletics department could mine the CRM data to identify that a particular season tick- The authors’ newly released book. et holder typically purchases nachos and a soft pretzel at the same concession stand at the end of the first quarter of every home football game. The analytics team would also have access to customer background information such as the birth date of each season ticket holder. Equipped with this data, a team representative could be waiting at the concession stand at the end of the first quarter during the football game that is closest to the ticket holder’s birthday in order to provide the ticket holder with a personalized thank you – and free nachos, soft pretzels and soft drinks for the entire family. An effective CRM data warehouse can also help sport organizations identify – and subsequently create “pitch packages” for – pre-qualified sales prospects [1]. Social media and digital marketing analytics: Gauging the value of social and digital media marketing campaigns “has become a large concern across the industry” [2]. Sport organizations attempt

W W W. I N F O R M S . O R G

to analyze both impression-based metrics (e.g., website page views, number of Twitter “followers” and similar key performance indicators) and attention-based metrics (e.g., measuring the authenticity, quality and extensiveness of consumer engagement) to determine the overall effectiveness of social and digital media marketing campaigns. Sport business industry leaders such as Bob Bowman (Major League Baseball president of business and media) understand that corporate sponsors “have gotten smarter about understanding that more subtle, immersive experiences on social media get

better results” [2]. For example, time spent watching video content on a website combined with relevant comments in response to a social media post are likely more reliable indicators of consumer engagement as compared with merely “liking” a Facebook post or visiting a website. Likewise, visual analytics applied to consumer Twitter posts of sport team or corporate partner logos/images might provide superior insight on the reach of (and consumer engagement with) a team or sponsor brand as compared with basic Twitter retweets and consumer use of hashtags [3]. ®

CAREER CENTER

Job Seekers: Find Your Next Career Move INFORMS Career Center contains the largest source of O.R. and analytics jobs in the world. It’s where

careercenter.informs.org JOB SEEKER BENEFITS • POST employers to you. • SEARCH • PERSONALIZED job alerts notify you of relevant job opportunities right to your inbox. • ASK the experts advice, resume critique and writing, career assessment test services and more!

www.informs.org | 800.446.3676

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

S PO RT S AN A LY T I C S Corporate partnership acquisition, valuation and evaluation: As explained by Mondello and Kamke [4], “One area of sport business research that continues to remain elusive centers on how to accurately quantify the respected return on investment (ROI) involving corporate sponsorships.” Industry research indicates “about one-third to onehalf of [United States] companies don’t have a system in place to measure sponsorship ROI comprehensively . . . many companies still do not effectively quantify the impact of these expenditures” [5]. Although sponsorships in the sport industry typically involve large financial investments, sponsors are “often at a loss in coming up with a viable means for measuring the ROI of these investments” [6]. Common sponsor objectives include the following: 1) improve brand reach, awareness and visibility via experiential marketing; 2) increase consumer brand loyalty and community goodwill; 3) drive retail traffic and showcase/sell product; 4) personalize client entertainment and prospecting; and 5) leverage the right to use a sport organization’s marks and logos (i.e., monetize intangible sponsorship assets). Evolving corporate partnership ROI and ROO metrics include the following measurement categories: 1) sponsor recall; (2) brand awareness, perception and affinity; 3) sponsor cost per consumer dollar spent (i.e., direct revenue from sponsor activation); 4) media impressions;

A N A LY T I C S - M A G A Z I N E . O R G

5) social media engagement; and 6) lead generation for future sales (See [7] for a detailed analysis of factors that influence sport sponsorship effectiveness). ❙ C. Keith Harrison is an associate professor within the College of Business Administration at the University of Central Florida, as well as an associate program director of the DeVos Sport Business Management Graduate Program. Scott Bukstein has been a faculty member at the University of Central Florida since 2010. Bukstein currently serves as the director of the Sport Business Management Undergraduate Program within the College of Business Administration at UCF. He is also an associate director of the DeVos Sport Business Management Graduate Program at UCF. Reprinted from “Sport Business Analytics,” edited by C. Keith Harrison and Scott Bukstein. ©Taylor & Francis, 2016. Reprinted with permission. REFERENCES 1. Smith, M., 2015, “Fan analytics movement reaching more colleges,” SportsBusiness Journal. 2. Spanberg, E., 2016, “Placing values on social media engagement,” SportsBusiness Journal. 3. Jensen, R. W., Limbu, Y. B., and Spong, Y., 2015, “Visual analytics of Twitter conversations about corporate sponsors of FC Barcelona and Juventus at the 2015 UEFA final,” International Journal of Sports Marketing and Sponsorship, Vol. 16, No. 4, pp. 3-9. 4. Mondello, M. and Kamke, C., 2014, “The introduction and application of sports analytics in professional sport organizations,” Journal of Applied Sport Management, Vol. 6, No. 2, pp. 1-12. 5. Jacobs, J., Jain, P., and Surana, K., 2014, “Is sports sponsorship worth it?” Retrieved from McKinsey & Company (http://www.mckinsey.com/ business-functions/marketing-and-sales/ourinsights/is-sports-sponsorship-worth-it). 6. Wolfe, M., 2016, “The elusive measurement dilemma of sports sponsorship ROI. Retrieved from http://www.bottomlineanalytics.com. 7. Kim, Y., Lee, H. W., Magnusen, M., and Kim. M., 2015, “Factors influencing sponsorship effectiveness: A meta-analytic review and research synthesis,” Journal of Sport Management, Vol. 29, No. 4, pp. 408-425.

W W W. I N F O R M S . O R G

ASSOCIATE CERTIFIED ANALYTICS PROFESSIONAL Analyze How aCAP Can Help Launch Your Analytics Career

The Associate Certified Analytics Professional (aCAP) is an entry-level analytics professional who is educated in the analytics process but may not have practice experience. This prestigious program begins a career pathway leading to the CAP designation.

www.certiďŹ edanalytics.org

CO N FERE N C E P R E V I E W

2016 Winter Simulation Conference The 2016 Winter Simulation Conference (WSC) has been the premier international forum for disseminating recent advances in the field of systems simulation for almost 50 years. The longest-running conference devoted to simulation as a discipline, this year’s WSC will be held on Dec. 11-14 at the Crystal Gateway in Arlington, Va., just outside of Washington, D.C. The The 2016 Winter Simulation Conference will be held Dec. 11-14 at the Crystal theme for WSC 2016 is “Simulating Gateway in Arlington, Va., just outside of Washington, D.C. Photo Courtesy of 123rf.com | tupungato Complex Service Systems.” WSC is designed for professionals and sustainability applications, and and academics from all backgrounds and general applications. Additionally, WSC across a broad range of interests. Aca- features introductory and advanced tudemic tracks include analysis methodol- torials, simulation education, case studogy, modeling methodology, simulation ies, Ph.D. colloquia, poster sessions, an optimization, agent-based simulation and extensive group of exhibitors and vendor hybrid simulations. Applied uses of simu- tutorials. Along with the lineup of tracks lation include social and behavioral simu- and speakers, WSC 2016 will also have lation, defense and security, modeling a special cross-fertilization track. and analysis of semiconductor manufacWSC 2016 is sponsored by technical turing, healthcare applications, logistics, co-sponsors ACM/SIGSIM, ASA, ASIM, manufacturing applications, networks IEEE/SMC and NIST, along with IIE, and communications, project manage- INFORMS-SIM and SCS. ment and construction, environmental Register now at www.wintersim.org. ❙

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

Learn how analytics and O.R. can maximize the value of your data to drive better business decisions.

http://meetings2.informs.org/analytics2016

http://meetings.informs.org/analytics2017

FIVE- M IN U T E A N A LYST

The ballot theorem In a closely contested election, it is possible that in any exit poll the eventual loser will lead the eventual winner due to simple random chance.

BY HARRISON SCHRAMM

This column will appear around the date of the U.S presidential election, and while I have been trying to avoid politics this year, it is simply impossible. I think it is fair to say that this year’s election has been contentious, and millions of people have closely followed the campaign coverage and debates. While in the United States votes are not reported until the election is over, there are “exit polls” on Election Day, which gain insight into the election by asking voters who they voted for as they leave their respective polling locations. In a closely contested election, it is possible that in any exit poll the eventual loser will lead the eventual winner due to simple random chance. Consider a simple non-election example. In a suite of cards (13 cards), there are nine number cards (2-10) and four non-number cards (Jack, Queen, King, Ace); correspondingly there are 36 number cards and 16 non-numbered cards in a deck. If one were to shuffle the deck, flip the cards over one at a time, and count the numbers as a vote for candidate A and non-numbers as a vote for candidate B, we can see that there are paths where the candidate B would be leading, even though A will be the eventual winner. Specifically, the probability that the first vote will show B having the lead on the first vote is 30 percent. Expanding this approach, counting paths, we can see that there are other paths, beginning with a vote

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

for A, which will have B in the lead, such as 9-K-Q and so on. The way to analyze this is to count paths, and there is an exceptionally elegant solution proposed by Bertrand [1], that the probability that the eventual winner is always ahead is given by: (p-q) (p+q) In our example for deck of Figure 1: Simulated heatmap of lead in a random shuffling of exit poll votes. As expected, the density of voting is darker nearer the fixed start and end points, and has maximum cards, there is only a 38 percent variability in the middle. This figure was produced using the Plotly library in R. chance that the “numbers” player will be ahead the entire time. between 100 and 150 persons per location. If for sake of example we take 100 voters in a particular location, APPLICATION TO THE 2016 PRESIDENTIAL ELECTION and use a current poll estimate, Mrs. Depending on the locale, it is either Clinton is leading Mr. Trump 55 percent illegal or discouraged to publish exit polls to 44 percent. While computing the before the polls in that state are closed. This probability that Mrs. Clinton leads all means that voters on the East Coast have day using the formula is trivial, we can no information as to how the voting actually use a simulation to consider the traces went, but voters on the West Coast (such during the day. as myself) will begin getting information A simulation is useful for seeing about how the nation voted before the polls this behavior in detail. While there’s in their state are closed. Voters may use a bit more code supporting the analythis information to influence their decision sis, the “business” of the simulation is: to vote themselves, depending on if they Library(magrittr) Votes = c(rep(1,55), see their candidate as winning by a large rep(-1, 44)) Votes %>% sample() %>% margin or losing by a small one. cumsum(). This is one of the (many) adThe Washington Post reports that vantages of using R Markdown as a doceach exit poll location attempts to poll ument preparation system.

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

FIVE- M IN U T E A N A LYST

is symmetric, so if you like Mr. Trump, the mathematics remain the same. A few technical issues to close this exploration: I have thought about random walks before in this article, particularly as applied to the Army-Navy football game, which has been the subject Figure 2: Surface plot of vote density in simulated exit poll. This graph was produced in R of the November column for using Plotly. the past two years. RanAs of mid-October (the time this was dom walks in the limit become Brownian written), the polls were changing so Motion, first explored in-depth by A. Einrapidly that it was an exercise in futility to stein in 1905. Our example here forms try to predict what the polls will be as we a special type, known as the Brownian approach Election Day. We will overcome Bridge. Unlike our previous examples, this by paramatrizing the possible leads, which started at a known starting point using the most historic landslide (35 and moved in either direction towards inpercent) as the benchmark. Note that this finity, the Bridge has a known start and end point. Further discussion is beyond the scope of this article, but is rich in theory and application. By the time this goes to press, and certainly by the time many of you will read it, the 2016 election will be history. ❙ Harrison Schramm (Harrison.schramm@gmail. com), CAP, PStat, is a principal operations research analyst at CANA Advisors, LLC. He is a member of INFORMS. REFERENCE Figure 3: Simulated lead changes in an election with 100 votes sampled and various lead sizes. As one would expect, the maximum number of lead changes occur when the race is even.

A N A LY T I C S - M A G A Z I N E . O R G

1. Grimmett and Stirzaker, 2001, “Probability and Random Processes,” Oxford University Press.

W W W. I N F O R M S . O R G

SPECIAL ADVERTISING SECTION

CLASSIFIEDS

MULTIPLE FACULTY POSITIONS Department of Electrical and Systems Engineering The School of Engineering and Applied Science at the University of Pennsylvania is growing its faculty by 33% over the next five years. As part of this initiative, the Department of Electrical and Systems Engineering is engaged in an aggressive, multi-year hiring effort for multiple tenure-track positions at all levels. Candidates must hold a Ph.D. in Electrical Engineering, Systems Engineering, or related area. The department seeks individuals with exceptional promise for, or proven record of, research achievement, who will take a position of international leadership in defining their field of study, and excel in undergraduate and graduate education. Leadership in cross-disciplinary and multi-disciplinary collaborations is of particular interest. We are interested in candidates in all areas that enhance our research strengths in 1. Nanodevices and nanosystems (nanoelectronics, MEMS/NEMS, power electronics, nanophotonics, integrated devices and systems at nanoscale), 2. Circuits and computer engineering (analog, RF, mm-wave, and digital circuits, emerging circuit design, computer engineering, IoT, embedded and cyber-physical systems), and 3. Information and decision systems (control, optimization, robotics, data science, network science, communications, information theory, signal processing, markets and social systems). Prospective candidates in all areas are strongly encouraged to address large-scale societal problems in energy, transportation, health, food and water, economic and financial networks, critical infrastructure, and national security. We are especially interested in candidates whose interests are aligned withthe schoolâ&#x20AC;&#x2122;s strategic plan(www.seas.upenn.edu/PennEngineering2020). Diversity candidates are strongly encouraged to apply. Interested persons should submit an online application at http://www.ese.upenn.edu/faculty-positions and include curriculum vitae, statement of research and teaching interests, and at least three references. Review of applications will begin on December 1, 2016. The University of Pennsylvania is an Equal Opportunity Employer. Minorities/Women/Individuals with Disabilities/Veterans are encouraged to apply.

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

CLASSIFIEDS

SPECIAL ADVERTISING SECTION

Assistant Professor

Department of Industrial Engineering University of Arkansas http://industrial-engineering.uark.edu The Department of Industrial Engineering at the University of Arkansas invites applications for a tenure-track Assistant Professor position with an anticipated start date of August 2017. We seek individuals whose research and graduate teaching interests in Analytics and other related areas that align with the department’s emphasis in the application of quantitative modeling and analysis in the areas of quality and reliability engineering, logistics and distribution, and healthcare systems. More importantly, we seek individuals who can make contributions to the university’s new cross-college, interdisciplinary Institute for Advanced Data Analytics. The new institute was established as a new industry research partnership for developing practical, implementable solutions to industry issues and problems as well as a source of continuing education in data analytics. The institute is led by close collaborations among the College of Engineering, the Sam M. Walton College of Business, and the Fulbright College of Arts and Sciences. Applicants should have a PhD in industrial engineering, operations research, statistics, computer science, or other closely related field and have excellent communication skills. Applicants should demonstrate potential for high-quality research, for securing competitive research funding and scholarly publications, provide evidence of teaching excellence (undergraduate and graduate courses), experience advising PhD students, and ability to provide appropriate service to the department, university, and the profession.

Faculty, Researcher, Postdoctoral Positions MIT Global SCALE Network The Massachusetts Institute of Technology’s Global Supply Chain and Logistics Excellence (MIT SCALE) Network is growing rapidly and we are seeking talented and energetic professionals for Faculty, Researcher, and Postdoctoral positions across the network. Candidates will help our centers become world leaders in education and research in supply chain management, freight transportation, global trade, and logistics. The MIT SCALE Network consists of six education and research centers of excellence focused in supply chain management and logistics. Currently there are centers in Cambridge, MA, USA; Zaragoza, Spain; Bogota, Colombia; Kuala Lumpur, Malaysia; University of Luxembourg, Luxembourg; and Ningbo, China. For more information and directions on submitting an application for any of these positions, please visit: https://academicjobsonline.org/ajo/SCM

The College of Engineering consists of eight departments that offer BS degrees in nine disciplines, MS degrees in ten disciplines, and a PhD degree in Engineering with several concentrations. The undergraduate enrollment in the college currently stands at over 3,300 and the graduate enrollment stands at almost 900. The college has 114 tenured/tenure-track full-time faculty associated with it and generally has externally funded annual research expenditures of approximately $20M. The college is a partner in a newly created NSF Engineering Research Center and multiple NSF I/UCRCs. Northwest Arkansas is one of the fastest growing areas in the nation having a population of over 500,000. Northwest Arkansas is home to the corporate headquarters of Fortune 500 companies Wal-Mart Stores, Tyson Foods, and J.B. Hunt Transport Services. Forbes Magazine has recently named Fayetteville, the home of the University of Arkansas, as one of the best places in the U.S. for business and careers. The Milken Institute has named Fayetteville the nation’s “Number One Performing City” and Livability.Com ranked Fayetteville as one of the top ten college towns in the country. Information about the area can be found at www.explorenwar.com. Applicants are asked to provide a letter of interest,curriculum vita, research and teaching statements,andthe names of three references. To ensure full consideration, application materials should be submitted online by December 1st, 2016 at http://jobs.uark.edu/postings/16265. Applications submitted after that date will be reviewed until the position is filled. Please direct any questions to: W. Art Chaovalitwongse, PhD Professor of Industrial Engineering 21st Century Research Leadership Chair in Engineering Co-Director of the Institute for Advanced Data Analytics 4207 Bell Engineering Center University of Arkansas Fayetteville, AR 72701 iesearch@uark.edu The University of Arkansas is an equal opportunity, affirmative action institution. The university welcomes applications without regard to age, race, gender (including pregnancy), national origin, disability, religion, marital or parental status, protected veteran status, military service, genetic information, sexual orientation or gender identity. Persons must have proof of legal authority to work in the United States on the first day of employment. All applicant information is subject to public disclosure under the Arkansas Freedom of Information Act.

A N A LY T I C S - M A G A Z I N E . O R G

The Department of Industrial Engineering at the University of Arkansas invites applications for a tenure track Assistant Professor position with an anticipated start date of August 2017. We seek individuals whose research and graduate teaching interests in Analytics and other related areas that align with the department’s emphasis in the application of quantitative modeling and analysis in the areas of quality and reliability engineering, logistics and distribution, and healthcare systems. More importantly, we seek individuals who can make contributions to the university’s new cross-college, interdisciplinary Institute for Advanced Data Analytics. The new institute was established as a new industry research partnership for developing practical, implementable solutions to industry issues and problems as well as a source of continuing education in data analytics. Applicants should have a PhD in industrial engineering, operations research, statistics, computer science, or other closely related field and have excellent communication skills. Applicants should demonstrate potential for high-quality research, for securing competitive research funding and scholarly publications, provide evidence of teaching excellence (undergraduate and graduate courses), experience advising PhD students, and ability to provide appropriate service to the department, university, and the profession. Applicants are asked to provide a letter of interest, curriculum vita, research and teaching statements, and the names of three references. To ensure full consideration, application materials should be submitted online by December 1st, 2016 at http://jobs.uark.edu/postings/16265. Applications submitted after that date will be reviewed until the position is filled. http://industrial-engineering.uark.edu The University of Arkansas is an equal opportunity, affirmative action institution. The university welcomes applications without regard to age, race, gender (including pregnancy), national origin, disability, religion, marital or parental status, protected veteran status, military service, genetic information, sexual orientation or gender identity. Persons must have proof of legal authority to work in the United States on the first day of employment. All applicant information is subject to public disclosure under the Arkansas Freedom of Information Act.

W W W. I N F O R M S . O R G

SPECIAL ADVERTISING SECTION

CLASSIFIEDS

APPLICATIONS INVITED FOR FACULTY POSITIONS IN INDUSTRIAL ENGINEERING AND MANAGEMENT SCIENCES The Industrial Engineering and Management Sciences Department at Northwestern University invites applications and nominations for two faculty positions beginning September, 2017. The positions are at the Assistant Professor or Associate Professor level. The searches are broad, with a preference for candidates in the following two areas: (1) computational statistics and (2) production, logistics, or healthcare. A strong commitment to rigorous and relevant research is essential.

Faculty Positions Ningbo Supply Chain Innovation Institute China Ningbo, China

We are pleased to announce the opening of the sixth center within MIT’s Global Supply Chain and Logistics Excellence (SCALE) Network in China.

The hires would have opportunities for interdisciplinary collaboration via broad University research initiatives in Optimization and Statistical Learning (www.osl.northwestern.edu), Engineering and Healthcare (www.ceh.northwestern.edu), and Transportation and Logistics (www.transportation.northwestern.edu).

The newly created Ningbo Supply Chain Innovation Institute China (NSIIC) will be joining the existing centers within the SCALE Network: the MIT Center for Transportation & Logistics or MIT CTL (Cambridge, MA, USA), the Zaragoza Logistics Center or ZLC (Zaragoza, Spain), the Center for Latin-America Logistics Innovation or CLI (Bogota, Colombia), the Malaysia Institute for Supply Chain Innovation or MISI (Shah Alam, Malaysia), and the Luxembourg Centre for Logistics or LCL (Luxembourg City, Luxembourg). Together, this network of centers educates hundreds of masters students, doctoral candidates, and executives each year.

The IEMS Department offers an undergraduate program, a Ph.D. program, a full-time professional master’s degree in analytics and a part-time professional master’s degree in engineering management. Both the undergraduate and graduate programs have been consistently ranked among the top ten by US News & World Report. Submit application electronically at www.mccormick.northwestern.edu/industrial/career/. Materials to be uploaded include a cover letter and a curriculum vitae detailing educational background, research and work experience. Applicants at the assistant professor level should also include one research paper and a statement of their current and future research program. Candidates will be asked to provide contact information for three references on the application site. To receive full consideration, all materials should be received by October 15, 2016; earlier application is encouraged.

We are recruiting for multiple faculty positions at all levels (Senior, Associate, and Assistant Professor) to support research activities in the area of transportation, logistics and supply chain management.

Chair, Faculty Recruiting Committee Department of Industrial Engineering and Management Sciences Northwestern University 2145 Sheridan Road, Room C210 Evanston, IL 60208-3119 facultysearch@iems.northwestern.edu

For more information and directions on submitting an application for these positions, please visit: https://academicjobsonline.org/ajo/SCM

Northwestern University is an Equal Opportunity, Affirmative Action Employer of all protected classes, including veterans and individuals with disabilities. Women, underrepresented racial and ethnic minorities, individuals with disabilities, and veterans are encouraged to apply. Hiring is contingent upon eligibility to work in the United States.

CAREER CENTER The INFORMS Career Center offers employers expanded opportunities to connect to qualified O.R. and analytics professionals, as well as a complete line of services to be used alone or in conjunction with the Career Fair at the 2016 Annual Meeting. Both give applicants and employers a convenient venue to connect. The Career Center is free to INFORMS members.

SEARCHING? HIRING? Access INFORMS CAREER CENTER • More analytics jobs • More O.R. jobs • Preferred jobs • Featured jobs • Job alerts • Anonymous career profiles • Powerful search capabilities

Nashville 2016

Learn more about the 2016 INFORMS Annual Meeting Career Fair, http://meetings.informs.org/nashville2016/career-fair.html

A NA L Y T I C S

N O V E M B E R / D E C E M B E R 2 016

THIN K IN G A N A LY T I CA LLY

Camping in the woods You have found yourself lost in the woods and you don’t expect to be found for several months. You’ll need to build a camp so you are in close proximity to water, food and firewood in order to survive the coming winter. Fortunately you have a map of the local area’s resources (Figure 1), which should help you decide where to set up your camp. The available resources are water (indicated by the blue drop icon), firewood (indicated by the brown log icon) and food (indicated by the red rabbit icon). Choose any green location to set up your campsite Figure 1: Where’s the optimal place to set up a campsite? with the goal of minimizing your daily travel distance to resources. Unfortunately some resources require multiple trips per day to meet your needs. Water requires three trips per day, firewood requires two trips per day, and food requires one trip per day. You must always return back to camp after visiting any resource location. Assume resources never run out. For example, if you build your camp at location I9, you will have to travel a total of 32 units per day (12 units for water because it is four units round trip times three trips per day, 20 units for firewood because it is 10 units round trip times two trips per day, and zero units for food because your camp occupies the same location). Question: At which location should you build your camp? Send your answer to puzzlor@gmail.com by Jan. 15, 2017. The winner, chosen randomly from correct answers, will receive a $25 Amazon Gift Card. Past questions and answers can be found at puzzlor.com. ❙

BY JOHN TOCZEK

John Toczek is the AVP Predictive Modeling at Chubb in the Decision Analytics and Predictive Modeling department. He earned his BSc. in chemical engineering at Drexel University (1996) and his MSc. in operations research from Virginia Commonwealth University (2005). He is a member of INFORMS.

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

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.