Analytics September/October 2017

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

SEPTEM BER/ O C TO BER 2017

Data Visualization The journey from machine-learning analysis to data-driven outcomes: presenting data in a powerful, credible way

ALSO INSIDE: • Dialing up call center productivity • Corporate profile: Air Liquide • Quantum computing, DNA & healthcare

Executive Edge Aspirent CEO Andrew Wells and Wunderman Data Mgmt.

VP Kathy Chiang on turning data into profits

INS IDE STO RY

Tech saves sabbatical Vijay Mehrotra, a professor in the business school at the University of San Francisco and the author of the popular “Analyze This!” column in Analytics magazine, recently returned from a seven-month sabbatical stay in Spain. Vijay took advantage of his time abroad to, among other things, try learning a new language, tour the country and soak up its culture, work on a book based in part on his Analytics mag columns and, fortunately for us, author new columns, the latest of which can be found on page 12.

As Vijay recounts in his column, the family’s Spain adventure got off to a rough start when the family dog somehow caught its foot in a moving sidewalk at the airport, resulting in an emergency trip to the vet, a three-day delay, rescheduled flights and assorted logistical problems that had to be solved in real time. Without giving the whole story away, suffice to say that Vijay, his family and the family dog made a nice recovery, thanks in large part to the “magic of 21st century technology.”

Vijay and I go back more than 20 years, when he was among the legion of high-tech start-up entrepreneurs plying their trade in the San Francisco Bay Area.

As I noted in a recent Analytics newsletter, at the time Vijay had an idea for a new

2

|

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

column for OR/MS Today, the membership magazine of INFORMS. Vijay described the column as “a cross between George Will, George Dantzig and George Carlin.

That is, a thoughtful (albeit partisan) look at operations research and management science delivered in a humorous tone bordering on the cynical.”

Given my admiration of the three Georges (especially Dantzig, the “father of linear programming”), I was intrigued.

When I received his first column, which appeared in the February 1998 issue of OR/MS Today, I was sold on Vijay’s unique ability to capture in easily understood yet powerful words the often inspiring, sometimes frustrating but always wonderful world of O.R. and analytics.

In 2010, shortly after INFORMS launched Analytics to reach out to the fastgrowing number of those with an interest in the word and the field it represented, Vijay jumped his column to the new mag, gave it a new name and described it as, “One man’s imperfect view on the rapidly evolving world of analytics.”

In a totally imperfect and sometimes outright crazy world, I find Vijay’s views make perfectly good sense. ❙

– PETER HORNER, EDITOR peter.horner@ mail.informs.org W W W. I N F O R M S . O R G

C O N T E N T S

DRIVING BETTER BUSINESS DECISIONS

SEPTEMBER/OCTOBER 2017 Brought to you by

38

FEATURES 38

VISUALIZING MACHINE-LEARNING ANALYSIS In the journey from analysis to data-driven outcomes, data visualization presents data in a powerful and credible way.

By Navneet Kesher

44

44

CALLING THE SMART WAY Big data analytics increase call center productivity and reduce unwanted phone calls by calling at the right time.

By Douglas A. Samuelson and Olcay Yucekaya

48

CORPORATE PROFILE: AIR LIQUIDE Multinational company’s Computational and Data Science R&D team supports an extensive, varied, dynamic research portfolio.

By Jeffrey E. Arbogast and Athanasios Kontopoulos

48 56

QUANTUM COMPUTING, DNA & HEALTHCARE Quantum computing based on DNA storage produces more accurate prediction of drug interaction with an individual’s genetic composition.

By Taran Volckhausen

56 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 and Simulation that Integrates with Excel

Everything in Predictive and Prescriptive Analytics Everywhere You Want, from Concept to Deployment.

functions; easy multiple parameterized simulations, decision trees, and a wide array of charts and graphs.

The Analytic Solver® suite makes the world’s best optimization software and 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.

Forecasting, Data Mining and Text Mining.

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.

Comprehensive Risk and Decision Analysis.

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

Analytic Solver is also a full-power, point-and-click tool for predictive analytics, from time series methods to classification and regression trees, neural networks, and access to SQL databases and Spark Big Data clusters.

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!

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

DRIVING BETTER BUSINESS DECISIONS

REGISTER FOR A FREE SUBSCRIPTION: http://analytics.informs.org INFORMS BOARD OF DIRECTORS

28

60

DEPARTMENTS

2 8 12 16 22 28 34 60 62 66 70

Inside Story Executive Edge Analyze This!

Healthcare Analytics INFORMS Initiatives Newsmakers Forum INFORMS Annual Meeting Winter Simulation Conference 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 ©2017 by the Institute for Operations Research and the Management Sciences. All rights reserved.

6

|

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

President Brian Denton, University of Michigan

President-Elect Nicholas Hall, Ohio State University

Past President Edward H. Kaplan, Yale University

Secretary Pinar Keskinocak, Georgia Tech

Treasurer Michael Fu, University of Maryland

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, Asia University

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, 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 Jeffrey 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., 1635 Old​41 Hwy, Suite 112-361, Kennesaw, GA 30152​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.: 678.293.5201

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.

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!

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

Ways to generate profit with the data you already have

BY ANDREW WELLS (LEFT) AND KATHY CHIANG

8

|

Build it and they will come. That is the view many organizations maintain about their data lakes and data warehouses. Companies are rapidly investing in systems and processes to retain business data that they know is valuable but have no clue what to do with it. Even the government collects mass amounts of data without specific plans for using the information at the time of collection. This trend only accelerates as the amount of data being produced continues to escalate. Today, it is estimated that human knowledge is doubling every 12 to 13 months, and IBM is estimating that with the build out of the “Internet of things,� knowledge will double every 12 hours.

Most organizations search for value in their data by throwing teams of data scientists at the various stores of data collected hoping to find insights that are commercially viable. This approach typically results in endless hours of digging for insights and if any

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

are found, they rarely see the light of day.

In order to monetize your data, you need a different approach, one that starts by turning the process on its head. We recommend three approaches to help you monetize your data: 1. It’s about the decision. A common approach when starting an analytics project is to ask what “questions” you would like the analysis to answer.

But if your goal is to drive actionable analytics that monetize your data, you need to start at a different point. You need to understand the “decisions” you would like the analytics to support. This approach, termed decision architecture, is radically different from conventional methods.

Understanding the decisions you would like to support drives the direction for the rest of the analytical exercise, including the type of data and analytics needed to support the decision. The decisions you focus on determine the analytics your team will undertake, which can range from simple metrics like ROI, or it may call for more sophisticated metrics such as a propensity or churn model.

2. Align decisions to business objectives. Knowing the goal is to provide analytics to support value driving “decisions,” you need to make sure the goals align with overall corporate objectives.

Through mapping your decisions to key A NA L Y T I C S

business drivers that achieve corporate objectives, you are charting a clear path to actionable analytics.

3. Economic value and decision theory. In order to monetize your data, adding economic value to your decisions through the use of data science and decision theory is a must. Whereas data science helps you generate insights from your data about actions you can take, decision theory helps you structure your decisions for maximum impact and feasibility. Economic value captures both the quantitative and qualitative aspects of an action and can come in various forms including revenue and profitability, market growth or process efficiency. The goal of economic value analysis is to provide the decision-maker with an understanding of the economic tradeoff among the set of decisions they have available to them.

Decision theory is applied to help decision-makers select the best choice to achieve their objectives. Structuring the decision criteria into a decision matrix laying out anticipated acts, events, outcomes and payoffs helps managers see more clearly the full scope of their proposed actions and make more objective choices, guarding against hidden or implicit cognitive biases. Cognitive biases arise where an individual holds S E P T E M B E R / O C T O B E R 2 017

|

9

EXE CU TIVE E D G E

economic value of your decision choices and employing decision theory to assist the decision-maker with making the best possible choice will improve the value of your decisions. These three practices will drive up the value of your analytics and enable you to monetize your data. ❙ Andrew Roman Wells is the CEO of Aspirent, a management-consulting firm focused on analytics.

Kathy Williams Chiang is VP, Business Insights, at Wunderman Data Management. They are the coauthors of “Monetizing Your Data: A Guide to Turning Data into Profit-Driving Strategies and Solutions.” For more information, visit www.monetizingyourdata.com.

Exhibit & Sponsorship information is now available.

http://meetings.informs.org/analytics2018

a view of a situation that is based on prior subjective experiences but may not be completely consistent with current reality. Confirmation bias, for example, occurs when the inclination is to look for information and analytics that support pre-existing beliefs or goals.

If you focus your analytics on your decision, you are already ahead of most analytical practitioners. Creating alignment from your decisions to your business drivers that achieve your corporate objectives makes your analytics actionable and relevant. Assessing

10

|

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

Big Data is only valuable if you know how to use it.

Now more than ever, data is everywhere.

Wouldn’t it be great to make sense of it all?

Bentley has been teaching analytics for 20 years, so we know data. With master of science degrees in Business Analytics and Marketing Analytics, you can customize a graduate program to fit your personal career goals.

To learn more about a graduate degree in analytics, visit bentley.edu/graduate

“FOR DATA SCIENCE, IT'S IMPERATIVE TO HAVE TARGETED TRAINING IN THE TECHNIQUES AND ALGORITHMS USED TO ANALYZE DATA.” - Amanda Dery, MSBA '17, Senior Data Scientist at Virgin Pulse

ANALY ZE TH I S !

Trip abroad reveals pros, cons of technology Just like that, a mere 45 minutes after leaving home, we found ourselves trapped in our own personal “no-fly” zone.

BY VIJAY MEHROTRA

12

|

Just in time for the beginning of the fall semester, my family and I have returned home from a sevenmonth overseas sabbatical, most of which was spent living in Madrid, Spain (a big thanks to my dear friend Javier Francisco Prieto for inviting me to visit the Department of Statistics at Universidad Carlos III de Madrid). It was by far my longest trip to Europe since my student days 30 years ago.

Our travels did not get off to a good start. After months of planning and preparations, our day of departure finally arrived in late January. Friends had driven us to the airport, where we checked in for our flight, got our boarding passes, went through security and walked toward our gate, cheerfully carrying our cabin luggage and walking our dog Spartacus.

Disaster struck suddenly and without warning.

At the end of a moving walkway, Spartacus abruptly began to whimper and squeal. Looking down, we realized that one of his little toes had gotten stuck under the metal at the end of the moving walkway.

His foot was pinned there for 30+ seconds before we could extricate it, at which point his toe was bleeding

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

somewhat heavily. My daughter’s instinctive and audible response – to repeat “Oh my god! Oh my god! Oh my god!” over and over again – gave voice to what we and literally hundreds of observers and bystanders were all feeling. It quickly became clear that the dog was going to the veterinarian, and that we would not be leaving for Spain that day as planned.

And just like that, a mere 45 minutes after leaving home, we found ourselves trapped in our own personal “no-fly” zone, unable to either return home (our house had been rented out to tenants) or to continue on toward our destination.

Ultimately, we were delayed – and dispirited – for three full days. The financial cost of the whole ordeal (vet bills, hotels, airline charges) turned out to be quite substantial.

There was nevertheless one silver lining amidst this darkness: As I scrambled to address a bunch of immediate logistical problems, I found myself marveling at the magic of 21st century technology. For starters, I was able to use WhatsApp on my smartphone to send a (free) text message to our landlord in Spain to inform her of our setback and to update her as new information became available. Our veterinarian e-mailed care instructions directly to a vet in our new neighborhood in Madrid (and to our air carrier, who required proof of the dog’s fitness to travel A NA L Y T I C S

before re-booking us). Our new boarding passes were issued electronically and scanned directly from our smartphones when we finally boarded our outbound flight. And once we landed in Madrid, our cabbie had no trouble finding our apartment in a narrow street in an old neighborhood thanks to a map app on his smartphone.

Almost as soon as we got out of the taxi, however, I began to curse our dependence on technology. The plan was to contact our landlord once we arrived, but in our exhausted state we realized that all of the contact information that we had for her was on our smartphones, which were no longer connected to any kind of network. From here, it was something of a comedy of errors, as I went first to the Irish pub on the corner (where they spoke English, took credit cards, but had no Wi-Fi network) and then to the bar across the street (where they had Wi-Fi, spoke no English, and didn’t take credit cards). Fortunately, by the time I returned from making these inquiries, our landlord had arrived to let us in (she had tracked our flight status online, of course).

Throughout our trip, I could not help marveling – and cursing – the role of technology in our travels, especially when contrasted with my days backpacking around Europe as a college student. Back then, I had relied on travelers checks and S E P T E M B E R / O C T O B E R 2 017

|

13

ANALY ZE TH I S !

cash transfers, and avoided restaurants that accepted credit cards (those were the expensive places at the time). This time around, we got cash from ATMs, and nearly every restaurant in Madrid cheerfully accepted our credit cards. Back then, we occasionally made (expensive) longdistance phone calls from (inconvenient) designated locations, whereas now we had many apps on our smartphones that let us talk to friends and family back home – and to cab drivers, Airbnb hosts and countless other people around the world – on a regular basis and for free! And back then, I would hungrily scarf down the occasional English-language newspaper or magazine. This time, I felt up-to-date at all times, thanks to the online versions of my favorite publications, not to mention Facebook and Twitter.

As we were winding down our time overseas, I received an e-mail from Harvard Business Review touting a series of articles on artificial intelligence and machine learning. In the lead article, Erik Brynjolfsson and Andrew McAfee argue that AI/ML should be thought of not as a collection of tools but rather as a general-purpose technology that will have far-reaching applications and a transformational economic impact.

Definitely worth a read.

Another piece in this HBR series, this one entitled “Inside Facebook’s AI Workshop,” caught my attention for a different 14

|

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

reason. Describing a conversation with Joaquin Candela, head of Facebook’s Applied Machine Learning group, author Scott Berinato observes that “AI is woven so intricately into the [Facebook] platform that it would be impossible to separate the product… from the algorithms.”

Probably because we had been overseas for several months when I read this quote, it immediately reminded me of how so many travel-related activities are now so tightly bound up with the enabling information technologies that they feel inseparable from one another. Thanks in large part to all of the IT that is already out there, technological change is coming at an even faster pace now, with AI/ML at the center of so much of it. As such, I anticipate that the transformation of travel is only getting started. Optimized driving routes based on real-time traffic data are here already, but this seems like just the beginning. On my next trip to Europe, I suspect that there will be a lot more intelligent algorithms embedded into the information technology that we travelers depend on.

Let’s hope it doesn’t take another 30 years for me to get back to find out. ❙ 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.

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

CPLEX Optimization Studio®.

Still the best and most complete optimizer and modeler Now you can get it direct

CPLEX Optimization Studio is well established as the leading, complete optimization software. For years it has proven effective for developing and deploying models and optimizing business decisions.

Now there’s a new way to get CPLEX – direct from the optimization industry experts.

Find out more at optimizationdirect.com The IBM logo and the IBM Member Business Partner mark are trademarks of International Business Machines Corporation, registered in many jurisdictions worldwide. *IBM ILOG CPLEX Optimization Studio is trademark of International Business Machines Corporation and used with permission.

HEALT H CARE A N A LY T I C S

A pragmatic approach Taming the challenges in healthcare with artificial intelligence.

BY RAJIB GHOSH

16

|

At last there is some respite for those of us in the healthcare analytics and technology business.

The neverending uncertainty and word storms emanating from the nation’s capital is now silent.

Perhaps temporarily but what a relief. We now can make strategies and plan for execution for the next six months to a year. In the meantime, if you would like to know about how the Affordable Care Act touched many lives for the better read this article in the Huffington Post. A Senate healthcare bill is now dead, and the hope of fixing what is not working in healthcare in a pragmatic bipartisan way is now rising. That’s the good news.

Now is a good time to focus on the future of healthcare analytics. Among my friends and colleagues in the industry, I have observed that the discussions in this area have somewhat moved from “big data analytics” to “artificial intelligence

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

(AI).� Analytics bolsters performance of AI. Machine learning makes AI become smarter. In my last couple of columns, I have described the visible euphoria among the innovator and investor communities, which has led a large number of startups to work on AI-based solutions for healthcare. So far, success has been limited, yet I love the enthusiasm; that’s what makes America great.

When hundreds of great minds start working on a problem everyone wins.

But before I go any further let’s do a quick recap of the state of the industry.

A recently published investment report from Silicon Valley Bank showed that techfocused investment firms are aggressively investing in healthcare companies that are developing artificial intelligence and machine learning technologies for biopharma and diagnosis tools (see Figure 1).

Another report from the healthcare startup accelerator Rock Health shows

Figure 1: Investment in healthcare companies that are developing AI and machine learning technologies for biopharma and diagnosis tools.

Source: Silicon Valley Bank

A NA L Y T I C S

S E P T E M B E R / O C T O B E R 2 017

|

17

HEALT H CARE A N A LY T I C S

that during the first half of the year 2017 investment in digital health technology companies remained strong with nine deals worth $100 million. None of those are AI companies. In February 2017, CB Insights reported that there are 106 AI startups in healthcare backed by various small, medium and large venture firms.

Not all will make it at the end, but it is

interesting to see how startups are now trying to offer “actions” rather than just platforms for “big data” analytics.

IBM WATSON: POSTER CHILD OF AI IN HEALTHCARE The million-dollar question is: How to create value using AI in healthcare?

IBM created its massive “Watson” expert

Figure 2: Investment in digital health technology companies remained strong in first half of 2017.

Source: Rock Health

18

|

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

system a few years back. The hope was Humans, essentially doctors, would have very high that this “supercomputer,” with to find time out of their already crammed almost bottomless knowledge and fast schedule to build the knowledge base for learning ability, would make doctors “irrelevant.” That, BECOME MORE AT THE however, has not happened . . . yet. The first few larger piFULLY ACCREDITED. lots failed to produce results.

The cost was high, just like NATIONALLY RANKED. the “big data” projects that usually work for large, deeppocketed health systems in the country. Costly deployments are meant to solve the most complex problems SS BUSINYETICS in healthcare such as develANAL oping precision medicine for cancer patients. That is a giant leap but not so pragmatic in the short term. Later NEW GRADUATE PROGRAMS on, IBM took Watson to the AND CERTIFICATES IN cloud and started offering its services in the “utility” business model for smaller-scale use cases and for smaller Get started at GRADUATE CERTIFICATE OR organizations. usd.edu/online So what is wrong in this picture about Watson, the so-called poster child for apWITH SPECIALIZATION IN plication of AI in healthcare?

BUSINESS Simply put, to make Watson ANALYTICS do its job, it needs certain cde@usd.edu types and volumes of data 800-233-7937 as the training data set.

Beacom School of Business

FLEXIBLE ONLINE LEARNING

BUSINESS M.B.A.

A NA L Y T I C S

S E P T E M B E R / O C T O B E R 2 017

|

19

HEALT H CARE A N A LY T I C S

the AI and connect the dots for Watson before the AI can perform on its own as an assistant. That’s proved to be a tall order for various reasons as evident from MD Anderson Cancer’s decision to walk away from the Watson project earlier this year.

Recently, the CEO of Social Capital venture fund referred to Watson as a “joke” in a CNBC interview. IBM later did a rebuttal; the technology is real, but clearly it is over hyped. PR and corporate marketing departments often get ahead of themselves, and there lies the problem. The AI industry needs focus, steady wins without undesired exuberance or marketing hype.

The technology needs to be robust, and it needs to solve real-world problems consistently at an affordable cost. Innovators need to understand the real problems, preferably firsthand. AI needs to create real business opportunities or unleash new tangible and quantifiable efficiencies.

PRAGMATIC USE OF AI IN HEALTHCARE: SHORT TERM AND LONG TERM In the long term, precision medicine for treating disease conditions is a very apt use case for AI. Watson, for example, can identify six different kinds of cancer. As the genomic data for a large population becomes more readily available, precise detection of gene mutation and corresponding medication (or gene therapy) can be identified by 20

|

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

AI in near real time. NIH has defined that as a “cancer moonshot” program with the objective to advance research for cancer therapies by 2020. In the short term, however, there are other use cases where first generation AI can bring value to the healthcare. A few of them are listed below: • Automated image analysis. Radiology is perhaps one of the most suitable fields for the application of AI. Image analysis is an established area for machine learning. IBM Watson is already being used for this use case by one imaging vendor. This area will see significant growth in the coming years.

• Intelligent patient triage and first level of primary care. Luckily, the ACA has survived for now. But owing to ACA, the demand for primary care is increasing. There is a shortage of primary care physicians.

While other transformational care delivery models using nurses instead of physicians are deployed in a limited way, an opportunity exists to deliver basic primary care via intelligent agents (AI) at retail clinic settings.

• Intelligent agents in mental healthcare. In a recent paper published by researchers from Stanford School of Medicine and Woebot Labs demonstrated how fully automated agents (i.e., AI) produces efficacy in the treatment of anxiety and depression among young adults (18-28). Given the nationwide shortage of behavioral health providers, this use case for AI could have major impact.

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

Overall, AI in healthcare still has a long way to go. Unlike consumer marketing or the advertising industry, healthcare is a difficult domain to penetrate. In healthcare, appropriate and adequate “signals” are harder to get to train AI quickly and continuously. I have no doubt that given the rate of exponential technology growth, a decade from now AI will become an integral part of the healthcare delivery. But for the time being I would like to share a dose of pragmatism with my colleagues who are getting inundated with the euphoria

CALL FOR ENTRIES A $15,000 Competition with a $10,000 First Prize

Application Deadline: October 27, 2017 KEY DATES FOR THE COMPETITION Friday, October 27, 2017 Deadline to provide a single pdf document containing a three-page summary of your achievement, and a cover page with a 60-word abstract, and the name, address, phone number, and affiliation of each author.

Monday, December 11, 2017 Finalists will be selected based on the summaries and the INFORMS Practice Section verification process.

Friday, February 9, 2018 Deadline for finalists to provide a full written paper.

Monday, April 16, 2018 Each finalist group will give an oral presentation of their work in a special session at the INFORMS Conference on Business Analytics & O.R. in Baltimore, Maryland, April 15−17, 2018.

A NA L Y T I C S

and PR campaigns coming from all directions about AI taking over healthcare soon. ❙ Rajib Ghosh (rghosh@hotmail.com) is an independent consultant and business advisor with 20 years of technology experience in various industry verticals where he had senior-level management roles in software engineering, program management, product management and business and strategy development. Ghosh spent a decade in the U.S. healthcare industry as part of a global ecosystem of medical device manufacturers, medical software companies and telehealth and telemedicine solution providers. He’s held senior positions at Hill-Rom, Solta Medical and Bosch Healthcare. His recent work interest includes public health and the field of IT-enabled sustainable healthcare delivery in the United States as well as emerging nations.

ABOUT THE COMPETITION

The purpose of the competition is to bring forward, recognize, and reward outstanding examples of operations research, the management sciences, and advanced analytics in practice. The client organization that uses the winning work receives a prize citation; the authors of the winning work receive a cash award.

ENTRY REQUIREMENTS

Visit the website www.informs.org/edelmanaward for detailed information.

Entries should report on a completed practical application and must describe results that had a significant, verifiable, and preferably quantifiable impact on the performance of the client organization. Finalist work will be published in the January-February 2019 issue of Interfaces.

Any work you have done in recent years is eligible, unless it has previously been described by a Franz Edelman Award finalist. Previous publications of the work does not disqualify it. Anyone is eligible for the competition except a member of the judging panel.

SUBMISSIONS INFORMATION

Please visit: www.informs.org/Recognizing-Excellence/2018-Edelman-Award for details on the competition and to submit your application.

S E P T E M B E R / O C T O B E R 2 017

|

21

INFO RM S IN I T I AT I VE S

Analytics workshop, Edelman competition, CAP benefits & more MANAGEMENT CONSULTANT, EDUCATOR TO LEAD ANALYTICS WORKSHOP Dr. Patrick Noonan (photo) will lead a two-day workshop, “Essential Practice Skills for High-Impact Analytics Projects,” on Sept. 26-27 at the Executive Conference Center in Crystal City, Va., where he will introduce a set of tools and techniques to help analytics professionals apply their skills to achieve high impact in the workplace. Presented by INFORMS, the workshop is part of the Institute’s “essential skills” continuing education program.

At the workshop, Noonan, a former management consultant, business owner and educator, will teach practical frameworks and systematic processes for addressing complex, real-world problems and how to facilitate effective action. Noonan combines a hands-on

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

teaching and learning approach honed from faculty posts at Harvard, Duke and most recently Emory University, with real-world experience from a business career spent at global management consulting giant McKinsey & Co, as well as Planning Technologies Group, which he co-founded in 1990. “Whether or not one directly handles data or models, they’ve become part of the nervous system of our society, and integral parts of the business models of most organizations we interact with,” Noonan says. “Some of the tools and techniques are obscured to us, but they’re there.”

Seats are limited, so register now at www.informs.org/continuinged. For more information, contact Bill Griffin at bgriffin@informs.org.

ENTER EDELMAN COMPETITION, SHARE YOUR BEST ADVANCED ANALYTICS WORK The Franz Edelman Award from INFORMS recognizes and rewards

A NA L Y T I C S

organizations for outstanding contributions of OR/MS and advanced analytics.

INFORMS encourages everyone whose work is driving impactful change and results to participate in the 2018 Edelman competition. To enter, submit a three-page summary of your most successful work and a 60-word abstract of the achievement by Oct. 27, 2017.

For more information about the most elite competition for the best advanced analytics or OR/MS project of the year, visit www.informs.org/2018edelman. For tips on helping INFORMS scout projects, click here.

THREE MORE REASONS TO HAVE CAP CERTIFICATION

By Jan paul Miller Recently, I wrote about three key benefits of certification: confirmation, differentiation and documentation. Here are three more benefits: productivity, networking and continuous professional development. Like the first group, these highlight the competence and capabilities of those who hold the credential.

I read once that employees with professional certifications are more productive because these credentials prepare workers for unexpected challenges in their jobs. A Certified Analytics Professional (CAP®) knows the seven domains associated with the CAP assessment

S E P T E M B E R / O C T O B E R 2 017

|

23

INFO RM S IN I T I AT I VE S

test, so they know how to create a process to solve a problem. Certified employees often work more efficiently than uncertified colleagues, and their presence can help improve the productivity of team projects.

Secondly, CAP certification opens the door to new networking opportunities.

Given that one’s certification shows competence in analytics, there is no other introduction needed. Whether among one’s peers in your place of employment or among certificants across the industry, it is a preferred calling card. Don’t leave home without it.

Finally, and very importantly, CAP certification commits the certificant to continuous professional development.

In my years in certification, I have seen any number of certificants in various professions complain about the requirements for re-certification. True, it involves a commitment of time and resources, but the benefits certainly outweigh those considerations. By gaining new knowledge, the CAP holder continues to make her/himself relevant by ongoing confirmation of currency in the knowledge, skills and attributes necessary for competent practice. I suspect 24

|

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

that something will have changed in the analytics ken by the time this article is published. As fast as the body of knowledge increases exponentially, it is critical for any certificant to keep current in the field.

CAP certification can be considered the highpoint in professional development, not only in displaying competence but being committed to lifelong learning as a credentialed professional. Taken together, these six benefits – confirmation, differentiation, documentation, networking, productivity and professional development – provide good reasons to seek and maintain CAP certification.

Jan paul Miller (jpmiller@informs.org) is the certification manager at INFORMS.

INFORMS NAMES INNISS DIRECTOR OF EDUCATION AND INDUSTRY OUTREACH INFORMS announced that Dr. Tasha R. Inniss has joined the INFORMS staff as the inaugural director of Education and Industry Outreach. In this role, she will contribute her extensive mathematics and education background to the overall vision, strategic direction and implementation of all education- and practice (industry)-related activities and outreach.

Prior to joining INFORMS, Inniss served as the acting deputy division W W W. I N F O R M S . O R G

director of the Division of Human Resource Development in the Directorate of Education and Human Resources at the National Science Foundation.

Concurrent with that, she was a tenured associate professor of mathematics at Spelman College in Atlanta.

Inniss graduated summa cum laude from Xavier University of Louisiana with a bachelor of science degree in mathematics. She earned a master’s degree

Ivan B. Class of ‘18 Oil and Gas

in applied mathematics from the Georgia Institute of Technology and a Ph.D. in applied mathematics from the University of Maryland, College Park. She was one of the first three African-American women to receive a doctorate in mathematics from UMD. Her first faculty position after completing her doctorate was as a Clare Boothe Luce Professor of Mathematics at Trinity College (now Trinity Washington University) in Washington, D.C. As an applied mathematician, her research interests are in the areas of data mining/data science, operations research

“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

S E P T E M B E R / O C T O B E R 2 017

|

25

INFO RM S IN I T I AT I VE S

and applied statistics. In addition to Inniss’ research interests, she also has a passion for teaching mathematics and encouraging undergraduate students to pursue degrees in STEM disciplines.

When she was a junior faculty member at Spelman College, Inniss received a Presidential Award for Excellence in Teaching because of her innovative teaching techniques and dedication to helping students to see how mathematics is applied in the real world.

“I am most excited about joining the dynamic INFORMS staff and working collaboratively to develop innovative integrated (education and practice) initiatives that help to support the goals and priorities of INFORMS,” Inniss says.

O.R. & ANALYTICS STUDENT TEAM COMPETITION The INFORMS O.R. & Analytics Student Team Competition challenges undergraduate and master’s level students to solve a real-world business problem using an O.R./analytics approach. Students from around the world can demonstrate their analytical skills and earn valuable prizes and prestige for themselves and their university.

Now in its second year, the contest provides student teams with a detailed problem statement, company data and ongoing access to company technical

26

|

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

and management staff. In addition, nine technology providers offer free access to their software for participating teams.

The eight finalist teams in the inaugural 2017 competition represented universities from the United Kingdom, Turkey, Singapore, Belgium and United States. “The competition allowed us to test our skills against brilliant students from around the world and also help make a lasting difference for the company,” one team leader said. “It was a truly rewarding experience. We look forward to building on what we’ve learned as we move forward in our studies and careers.”

Principal, a leading financial investment management company and 2018 title sponsor, will provide the 2018 challenge problem. The company is looking for insights and innovative ideas on developing a global equity optimization framework to improve risk adjusted returns. In addition to in-depth background information and proprietary data sets, Principal will offer students access to company staff through a unique “office hours” program, webinars and Q&A opportunities.

Interested students should subscribe now for updates and deadlines (http:// connect.informs.org/oratc/home).

The preliminary problem statement was published on Sept. 1, and the final problem, data sets and software available will be announced on Oct. 13. Beginning W W W. I N F O R M S . O R G

Oct. 13, teams can register their intent to compete. While registration is nonbinding, it indicates a team’s interest and allows access to the final problem and data. In addition, only registered teams will be invited to informational webinars, Q&A opportunities and “office hour” sessions with Principal staff being held in November-December. Written entries are due Jan. 29, 2018.

A panel of industry and academic experts will judge written submissions based on teams’ use of the full analytics process, from framing the problem to

methodology selection, data use, model building and quantitative analysis. Demonstrated “soft skills” in teamwork, communication and presentation will also be considered in judging.

Eight finalist teams will be announced on March 5, 2018, and invited to present their work at the INFORMS Analytics Conference on April 15-17, 2018, in Baltimore. Finalists receive a travel stipend and complimentary registration to the conference. For complete information, visit: http://connect.informs. org/oratc/home. ❙

Apply for this INFORMS competition that provides workplace experience for undergraduate and master’s level students. Teams use O.R. and Analytics to make decisions and solve a real-world business problem.

Finalists will present solutions live to an expert panel at the 2018 INFORMS Conference on Business Analytics and Operations Research in Baltimore, Maryland.

Monetary awards will be granted to all finalist teams, as well as award certificates for each team member and school.

Finalists will also receive a travel stipend and complimentary registration to the Analytics Conference.

TITLE SPONSOR:

RE! /home O M ratc

RNrms.org/o A E L .info

h

ct

nne

/co ttp:/

PRESENTING SPONSOR:

Founding Sponsor

A NA L Y T I C S

S E P T E M B E R / O C T O B E R 2 017

|

27

NE W S M AK E R S

Small retail formats, cognitive computing, AI and deep learning GAINING DISTRIBUTION IN SMALL RETAIL FORMATS BRINGS BIG PAYOFFS

Photo Courtesy of 123rf.com | © radub85

Small retail formats with limited assortments such as Save-A-Lot and Aldi and neighborhood stores like Target Express have been growing in popularity in the United States and around the world. For brands, the limited assortments mean greater competition for shelf-space, raising the question of whether it is worth expending marketing effort and slotting allowances to get on to their shelves.

According to a forthcoming study in a leading INFORMS scholarly marketing journal, Marketing Science, the answer is “yes.”

Gaining the same increase in distribution penetration from limited assortment stores provides much larger increases in sales, relative to that from large assortment stores.

The study, “The Effect of Retail Distribution on Sales of Alcoholic Beverages,” is co-authored by Richard Friberg and Mark Sanctuary 28

|

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

of the Stockholm School of Economics. In general, answering the question of how much distribution affects sales is challenging, due to a “chicken and egg” problem. Sales increase with distribution, but firms also increase distribution when sales increase, making it hard to tease out the effect of distribution on sales from the effect of sales on distribution.

The authors examined SKU level sales data from Systembolaget, a Swedish state-owned monopoly for retailing alcoholic beverages such as wine, beer and other spirits from 2006-2011. The advantage of the Swedish data is that the monopoly has to follow a certain set of rules when changing distribution for brands. These rules were established to ensure fairness to wholesalers and to satisfy conditions for Sweden’s accession to the European Union. The Swedish retailer classified stores into four tiers based on assortment levels. All brands chosen for distribution are initially sold at the largest format stores. As they become more popular, their likelihood of being sold in the next tier of smaller format stores rises. These distribution decisions are all made twice a year.

Friberg explained, “Given we knew the firm’s rules and timing for distribution changes, we know which brands have similar chance of being chosen for distribution in the next tier at a given time. From A NA L Y T I C S

among this set of similar brands, we compare the increase in sales for those who gained distribution to the next tier relative to those that did not. This allowed us to isolate the effect of distribution on sales, because the rule allowed us to eliminate the effect of sales on distribution.”

The authors found that a 10 percent widening in the retail distribution of wines, for example, yields increases in sales by 1.2 percent, 2.1 percent and 6.2 percent as the product gained distribution to the next tier of stores with smaller assortments. Sanctuary noted, “Our results suggest that gaining distribution in one large store with a certain turnover is less effective in increasing sales than gaining distribution at two small stores that have the same combined turnover as the larger store. Hence gaining distribution in small retail formats remains important for larger brands.”

The authors dig deeper into why they find that the effects of increasing distribution to smaller format stores are greater.

They rule out explanations such as greater word of mouth that arises from larger distribution and conclude that a brand gaining distribution in limited assortment stores can get a larger share of sales relative to shares at larger assortment stores. “Getting a larger share of a smaller pie at limited assortment stores is well worth fighting for,” said Friberg, with clear implications for S E P T E M B E R / O C T O B E R 2 017

|

29

NE W S M AK E R S

brands deciding to expend effort in gaining distribution at such stores that are growing around the world.

COGNITIVE COMPUTING A DISRUPTIVE FORCE, BUT ARE CMOs READY? While marketing and sales professionals increasingly find themselves drowning in data, a new IBM study finds that 64 percent of surveyed CMOs and sales leaders believe their industries will be ready to adopt cognitive technologies in the next three years. However, despite

this stated readiness, the study finds that only 24 percent of those surveyed believe they have strategy in place to implement these technologies today.

Marketing and sales outperformers are ahead of the cognitive game with 73 percent already collecting and analyzing external market data. Surveyed executives from businesses that have outperformed their competition for the past three years in revenue growth, profitability or other factors made up 13 percent of the study. Of these surveyed outperformers, 93 percent believe cognitive computing is

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 5,500+ attendees open and read eNews Daily.

DELIVERY SCHEDULE:

ADVERTISING RATES:

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

Any 1 day................ $250

Wed., 10/18 ..............preview edition to all meeting pre-registrants

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

Sun., 10/22 .............. all attendees

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

Any 2 days.............. $450 Any 4 days............... $700

Mon., 10/23 ............. all attendees Tues., 10/24 ............. all attendees Wed., 10/25 ............. all attendees

QUESTIONS: Olivia Schmitz, Exhibit & Sponsorship Sales Manager Phone: 443-757-3539 email: olivia.schmitz@informs.org EXHIBITORS & SPONSORS RECEIVE A 10% DISCOUNT.

30

|

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

An

nu M eN e a EW e t l SD i aily n g W W W. I N F O R M S . O R G

A NA L Y T I C S

Photo Courtesy of 123rf.com | © Natalia Romanova

mature and market ready, and 91 percent assert that cognitive computing is good for their organizations.

According to the study, “From data deluge to intelligent insights: Adopting cognitive computing to unlock value for marketing and sales,” while both CMOs and heads of sales agreed that “customer satisfaction” is the number one value driver for adopting cognitive solutions, practically speaking, many of those surveyed say they aren’t sure their organizations are currently set up to make a successful transition.

The study, conducted in cooperation with Oxford Economics, is based on a global survey with 525 CMOs and 389 heads of sales across industries to determine the extent by which marketers and sellers aim to embrace cognitive.

Cognitive computing, such as IBM Watson, is a next-generation technology that can quickly understand and reason vast amounts of structured and unstructured data, like sounds and images, in the same way humans do – by reasoning, learning and interacting to improve accuracy overtime. While traditional analytics can provide data for businesses to draw insights from, cognitive can more easily predict outcomes and turns those

insights into actionable recommendations, which can impact real business decisions.

For surveyed CMOs, they expect the real advantage of cognitive lies in two key areas: improved customer experience and financial results, including increased financial yields and improved ability to identify marketing ROI. For sales leaders in the study, it’s all about finally achieving a 360-degree understanding of customers so they may better predict their customers’ needs and improve prospecting, lead strategy, customer service and experience. For example, HSN is using cognitive to help its stories reach the right audience

S E P T E M B E R / O C T O B E R 2 017

|

31

NE W S M AK E R S

on its preferred channel, which encourages more viewers to become customers and drives HSN’s business growth.

WHY HUMAN EMOTIONS WILL FUEL THINKING (AND SUCCESS)

Photo Courtesy of 123rf.com | © Prasit Rodphan

We are on the cusp of a new era, led by artificial intelligence and deep learning. This so-called Smart Machine Age will lead to technology and robots outperforming humans in many tasks. This is bad news on the job front. In fact, research from the University of Oxford states that almost half of jobs in the United States will be automated over the next 15 years. Further, based on that research along with independent research, the chief economist of the Bank of England predicted that the United States could lose upward of 80 million jobs during that time frame.

At first glance you might think the solution is for you to become more robot-like to fit into this brave new high-tech work world.

32

|

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

Ed Hess disagrees. He says the key to staying employable in the Smart Machine Age is to further excel at what makes us unique as human beings – our real, not artificial, emotional and social intelligence.

“The technology of the future, led by artificial intelligence and deep learning, will be able to outthink us in many ways,” says Hess, co-author along with Katherine Ludwig of “Humility Is the New Smart: Rethinking Human Excellence in the Smart Machine Age.” However, in the coming Smart Machine Age, Hess says that “our emotional intelligence will be the very factor that makes us unique and employable.”

Hess and Ludwig explain that our emotions fuel our imaginations and enable personal connections to others in ways that machines cannot replicate. When we use our emotions to serve and collaborate with each other, there are no limits to our thinking, creativity and performance.

“In the Smart Machine Age, you do not want to behave like a machine,” Ludwig says. “As smart machines take over more jobs, the most successful people will be those who can leverage their emotions and the best of their humanness to think better and be more creative, innovative and collaborative.” ❙ W W W. I N F O R M S . O R G

GREAT STORIES CREATE LASTING IMPRESSIONS...

YOURS SHOULD TOO.

Learn how to become a compelling data storyteller.

Register today for ESSENTIAL PRACTICE SKILLS FOR HIGH-IMPACT ANALYTICS PROJECTS AMA Washington, DC Area Executive Conference Center 2345 Crystal Drive, 2nd floor Arlington, Virginia 22202 September 26–27, 2017 8:30am-4:30pm

Register at www.informs.org/continuinged

FO RUM

Bias detection crucial to today’s decision-making How new programs and apps sniff out social media bias and ‘fake news’ using sentiment analysis.

BY CAROL OZEMHOYA

34

|

The big buzz word phrase nowadays is “fake news,” and it seems people are more confused than ever on what news sources to trust, whether it’s TV news broadcasts (local and network), newspapers or the Internet, especially social media.

Unfortunately, some unscrupulous players have learned how to use all of these news sources to scramble the truth, as well as straight out lie to throw the public on a course of their choice, whether it hurts the economy, the public perception of politicians, politics and high-profile situations or not. What people, including financiers, regular “Joes” and “Jills,” and business executives may not realize is that there are now programs and apps available that can help citizens detect media bias as well as so-called fake news items.

Bias detection can be assessed through sentiment analysis (a form of artificial intelligence), which uses social media and news stories to determine just how “offbeat” something – a story or a

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

post – can be. These highly accurate programs can detect the attitude of the elements in the story or post (including the author), as well as the type of reaction – emotional and otherwise – which the post is hoping to convey.

The attitude may be a judgment or evaluation of something or a person, affective state (the emotional state of the author or the speaker) or the intended emotional communication (the intended emotional affect). Millennials seem to especially be the target of these posts, as they make up a huge voting block and represent huge financial clout.

DETECTING SENTIMENT HAS BECOME INCREASINGLY IMPORTANT There are companies that have diligently developed some tools to assist in detecting media bias. One of the things this involves is “opinion mining.”

According to Bizcommunity.com, this is “the science of using text analysis to understand the drivers behind public sentiment.”

In today’s high-stakes world, this is becoming increasingly important for executives that handle financial decisions for those just investing mere thousands as well as those with millions and even billions to put at risk (let’s be real – there is really no such investment that is a “sure thing”).

A NA L Y T I C S

Why is it important to gather sentiments (including emotions and opinions)? “Gather enough opinions – and analyze them correctly – and you’ve got an accurate gauge of the feelings of the silent majority,” says JP Kloppers of Bizcommunity. This relates not only to how people feel, “but the drivers underlying why they feel the way they do.”

The mined data can be used to help a company pinpoint how it’s doing in the global marketplace, or a financial analyst can use it to gauge more closely what the immediate future holds for a stock. This, of course, can translate into millions of dollars of profit and even to stymie losses.

It’s a huge tool for those smart enough – actually hip enough – to put to use in their everyday operations, especially at analysis and management levels. That applies to political movements, as well as simple customer service agendas for fast-food joints. Detecting media bias will become the “in” tool on a massive scale for anyone determined to make a campaign or product a resounding success.

HOW IT’S DONE The arts of artificial intelligence are developing rapidly among some key players in this highly concentrated world. Mining data (including media stories and posts) and analyzing it has become the focal S E P T E M B E R / O C T O B E R 2 017

|

35

FO RUM

point, for example, of Houston-based Indexer. The firm has an API (application program interface) that is designed to identify the current state of media opinion on specific subject matter as expressed as a cross-sectional visualization of sentiment polarity (emotion) and subjectivity (opinion) associated with the named entities inside of the news articles people read.

In other words, researchers can apply specific programs to understand the “fair and balanced” nature of the news content chosen to be reviewed. In fact, researchers can answer the question: Are

media sources such as CNN, Fox News and MSNBC biased news organizations?

It may sound complicated, but the mechanisms being developed and those already available can feed decisionmakers and analysts more accurate information, and in the long run, save a firm from heavy losses or help create new gains in marketplaces around the world, big and small. ❙ Carol Ozemhoya is a writer at Indexer, a tech startup in the artificial intelligence space with a focus on computer vision and natural language processing technologies.

Connecting with the right audience isn’t rocket science.

Exhibit & SPonsor at the world’s largest Operations Research & Analytics Conference For additional details, contact Olivia Schmitz at Olivia.Schmitz@informs.org or visit: http://meetings.informs.org/houston2017 Houston, Texas | October 22-25, 2017

36

|

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

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

DATA VIS UA LI Z AT I O N

Output visualization of machine learning analysis

BY NAVNEET KESHER ata science is more than just building machine learning models; it’s also about explaining the models and using them to drive data-driven decisions.

In the journey from analysis to data-driven outcomes, data visualization plays a very important role of presenting data in a powerful and credible way.

is unstructured data – texts, blogs, documents, photos, videos, etc. Unstructured data, also known as dark data, includes information assets that organizations collect, process and store during regular business activities, but generally fail to use for other purposes (for example, analytics, business relationships and direct monetizing). Unstructured data is the hidden part of the massive iceberg that has yet to WHY UNSTRUCTURED DATA? be analyzed for useful decision-making.

Structured data only accounts for about In many circles, unstructured data 20 percent of stored information. The rest is considered a burden that should be

D

38

|

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

Figure 1: Unstructured data: the hidden part of the massive iceberg.

sorted and stored away. In reality, it contains valuable business insights that can significantly augment the business understanding that we have today from structured data.

Although machine learning can analyze any type of data (structured or unstructured), unstructured data is virtually useless without machine learning algorithms (including natural language processing (NLP) algorithms, text-mining algorithms, pattern/classification algorithms, etc.) While machine learning algorithms have seen significant advancements, the tools and processes to visualize the results from these algorithms for the common man have not kept pace.

A NA L Y T I C S

Visualization tools for unstructured data are extremely valuable, but they have traditionally operated mostly on highly structured data, such as stock prices and sales records. As we create and consume more unstructured data, we have to extend the visualization efforts to include unstructured data.

IMPORTANCE OF DATA VISUALIZATION As a data scientist, I always question the amount of time I put into data visualization. Throughout my early analytics career, I observed that the prettier my graph, the more skeptical my audience was in the quality of my analysis. While I loved data

S E P T E M B E R / O C T O B E R 2 017

|

39

DATA VIS UA LI Z AT I O N

REPRESENTING UNSTRUCTURED DATA Below are the three broad guidelines that I follow while building visualization for unstructured data: 1. Start with a goal.

Goals are the fundamental bonding agent that connect the purpose of the analysis to the visualization of results. Whether the goal is to arrive to a decision or start an action into Figure 2: Can pretty graphs make an audience more skeptical of the quality of the analysis? exploring next steps, the visualization, I also feared coming out data scientist should aim to identify and as the person who puts more emphasis convey the results and corresponding and effort into making the graphs visualization that best supports a wellpretty rather than ensuring a thorough defined goal. analysis (Figure 2).

For example, if the goal is to analyze As I progressed in my career, I a call center’s audio recordings to deterrealized that data analysis and data mine the type and corresponding volume visualization are not entirely exclusive of complaints, a cubism horizongraph work sets – they co-exist and feed off of may be very useful. Cubism.js is a very each other (aka, you can produce pretty effective time series visualization tool graphs and still come off as someone that uses stacked area graphs to help with analytical prowess). The rest of analyze output content from audio-video this article on data visualization will streaming data. In the case of call center focus on representing highly complex audio recordings, the horizongraph visuanalysis on a sheet of paper (or slide) alization can help determine the intensity for someone who may not have the of the customer conversations (along need to understand the underlying with time series data) without having to details. transcribe audio into text.

40

|

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

Figure 3: Visualization is most effective when it is simple to understand and can stand by itself.

While we are on the topic of call center’s customer service recordings, if the goal is to understand the differences between a subscription customer vs. a free-tier customer, then a text analysis along with scatter text visualization may make more sense. Of course, this will need transcription and annotation of the media files.

Call centers use analytics for analyzing thousands (or millions) of hours of recorded calls. Among others, the main goal is to gain insight into customer behavior and identify product/service issues. The analysis method that I have found particularly useful for these goals A NA L Y T I C S

is self-organizing maps (SOM), which, along with classification, have added the benefit of dimensionality reduction.

SOMs are also good for visualizing multidimensional data into 2-D planar diffusion map.

Having and understanding the goal is the most crucial aspect for any data visualization process. Always ask yourself and your stakeholders: What will this data be used for? List the data points that will be vital for answering strategic questions for your business and then create a wireframe of the story that is going to engage your audience.

S E P T E M B E R / O C T O B E R 2 017

|

41

DATA VIS UA LI Z AT I O N

2. Simplicity for the win. The very reason we analyze unstructured data is to provide structure to it. Data visualization plays a very important role in conveying the results of the analysis, and visualization is most effective when it is simple to understand and can stand by itself without a lot of subtext or metadata. One of my favorite examples is this visualization on “How Families Interact on Facebook” by the Facebook Data Science Team.

This is a very simple, yet powerful way to reveal the results of a very complex text analysis.

Another classic example is the use of bar/line charts vs. radar/spider charts.

I am a big proponent of easy-to-read charts, aka, charts that can convey a maximum amount of information in the least amount of time.

Here are a couple of other ways I like for simple visualization of unstructured data: Word clouds. Word clouds help visualize the occurrence of words within a

Figure 4: Classic example bar/line charts vs. radar/spider charts.

42

|

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

corpus, with the size of the text representing the number of times the word or the phrase occurs in the larger text collection.

Word clouds are very effective in visualizing the results when performing TF-IDF (term frequency–inverse document frequency, a numerical statistic that is intended to reflect how important a word is to a document in a collection or corpus of words). Word clouds can be very effective in uncovering the topic areas of discussion for any social media content or feedback surveys/comments. If you use Python, you may want to bookmark this awesome word_cloud library by Andreas Müller. An interesting application of TFIDF with word clouds can be found here.

Chord diagrams. A chord diagram is a powerful tool that can be used to represent the contextual meaning of words (especially when analyzing using latent semantic analysis). If the number of topics are <10, Seaborn heatmaps (or even a stacked bar chart) may be a better alternative; however, with a larger set of topics, chord diagrams have better visual representation.

Click on the following for Python examples of Chord Diagram and Filled Chord Diagram. The lines connecting the nodes on the circle indicate that the W W W. I N F O R M S . O R G

relationship between these nodes/words (color of the line can denote a positive or negative relation) and the thickness of the connecting lines quantifies the extent of the relationship.

Don’t overload your visualization with data and present clear contrasts wherever applicable.

3. Know your audience. Knowing your audience and tailoring visualization for optimal consumption will go a long way into making of a successful presentation.

It’s always good to understand how the data translates into strategic direction for the product. Don’t work in a silo – involve and get feedback from your stakeholders as you do your analysis and create visualization thereof. Iterate! If you cannot get feedback from everyone, make sure that you think about who will actually be looking at these visualizations, what’s important to them and most importantly, how much time will they have to look at your graphs. These are the most important things you should do to understand your audience. Technical jargon won’t work if your audience doesn’t know what they mean. No matter how beautiful your graphs are, if you don’t deliver meaningful and actionable insights, your work does not classify as impact.

For example, if you present your data in form of a network graph (network A NA L Y T I C S

Figure 5: The four quadrants of good analytics insights.

graphs are designed to measure and quantify the relationships between different vertices or nodes on a graph), take some time to explain how the graph works. In a social data context, network graphs can be a powerful tool in telling a story on the health of your product’s ecosystem.

Data visualization is an art that data scientists need to be good at in order to tell a compelling story from their analysis.

Figure 5 best depicts the four quadrants of good analytics insights. ❙ Navneet Kesher (navneet.kesher@gmail.com) is the head of Platform Data Sciences at Facebook.

Prior to joining Facebook, he served as a manager of analytics for Amazon. Based in the Greater Seattle Area, he holds an MBA from the University of Southern California.

S E P T E M B E R / O C T O B E R 2 017

|

43

N E W T E CH NO LO GY

Calling the smart way Big data analytics increase call center productivity and reduce unwanted phone calls.

BY DOUGLAS A. SAMUELSON (left) AND OLCAY YUCEKAYA

IF

we know anything about outbound telephone call centers, itâ&#x20AC;&#x2122;s these two facts: 1) People profess to hate all the unwanted telephone calls, and 2) Call centers keep making the calls because many of them actually do produce sales, debt payments and votes in elections. Now new research and development has produced a better approach, increasing response while decreasing the number of calls.

This seemingly paradoxical improvement reflects a profound insight: Most call centers have focused on increasing the number of dialing attempts, while trying to limit calling when no representative will be available to talk to the answering party.

In fact, doing this much more effectively

44

|

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

revolutionized the call center industry back in the late 1980s. The method resulted in a major patent [1] and a Wagner Prize finalist entry [2], as well as great success for the company that developed it.

Recently, 2Contact, a call and contact center company in Haarlem, the Netherlands, realized that a different focus would work better â&#x20AC;&#x201C; not more calls, but calls better timed to reach the right party when the person is most amenable to being approached. This means that the software, trade named Smart-Call (tm), uses big data and current high-speed computing capabilities to schedule call attempts, based on past responses by the called person or similar people. Neither the data nor the high-speed computations were

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

Better-timed calls reach the right party when the person is most amenable to being approached.

available in the 1980s, but that was then, and this is now.

Ironically, this was initially a tough sale within 2Contact. Olcay Yucekaya, the inventor of Smart-Call, says, “It may sound hard to believe, but I’ve had more failures and rejections than you can imagine. In the end, ignoring the rejections and always thinking in solutions resulted in the fact that we are the leading call center in this idea.”

In fact, many U.S. companies have tried and are trying to integrate better data about called parties into their dialing systems. Since most of these approaches are closely guarded trade secrets,

A NA L Y T I C S

companies that claim to do superior list management are reluctant to disclose few if any substantiating data. We surmise, however, that these newer methods still rely on a predictive dialing method like the 1986 original, possibly modified to take advantage of cloud computing and real-time massively parallel simulations [3]. Such systems can adapt more quickly and smoothly to changing conditions (there’s no steady state in outbound calling, so traditional queuing models don’t help – it’s an adaptive automatic control problem, and rapid smooth adaptation is critical), but they still lack the shift in focus of Smart-Call.

S E P T E M B E R / O C T O B E R 2 017

|

45

CAL L C EN T E R P ROD U C T I V I T Y

In tests using closely matched pairs of lists, 2Contact has observed that the list using the new method had a substantially higher rate of sales – 31 percent more, in the most rigorous test tried. The test group had a 4 percent higher answer rate – that is, somewhat more people were willing simply to answer the phone and talk to the representative if they were called at a more convenient time. But the difference in sales was much larger, indicating that the effect of time of call was much more than just whether people would answer.

Another performance statistic is the proportion of records in a contact list eventually used in the campaign. Most campaigns have a preset limit of number of times a party will be called. If this number of calls does not result in a conversation, the record is just abandoned for the purpose of a particular campaign. Usage rates around 60 percent to 70 percent are common; with a well-targeted campaign using older methods, it is unusually good if as many as 75 percent of the records end up being used. In the matched groups test, the “standard” group has 69 percent usage – but the test group had 93 percent.

And what about unwanted calls?

The new method substantially reduces the number of call attempts per hour per representative, and even the number of 46

|

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

completed calls per hour per representative. In the matched groups test, the test group averaged 4.31 completed calls per representative per hour, versus 5.54 for the other group; but the test group had 3.54 effective calls per hour, that is, calls that result in either a buying decision or a no-buy decision, versus 3.34 for the other group. (Completed calls also include connections to people who do not stay on the phone long enough to have the intended conversation, plus call sequences that reach the set limit on number of call attempts.)

Not surprisingly, the representatives in the test group had longer talk times, averaging three minutes and 19 seconds as compared with two minutes and 59 seconds for the other group. Conversations with the right party, when the person is receptive, do tend to last longer.

Also quite noteworthy is the abandoned call rate, the proportion of call attempts that result in the system hanging up on the called party because the dialing went too fast and there was no representative available to talk when the called party answered. The new method cut this rate almost in half, from 1.9 percent of attempts for the other group to 1.0 percent for the test group.

The higher proportions of answers and long connects with the new targeting approach makes the system more stable W W W. I N F O R M S . O R G

and enables less “aggressive” dialing to keep representatives busy. The test group had an average of 24.38 dialing attempts per representative per hour, versus 30.85 for the other group. In short, Smart-Call reduces the call center’s costs and the number of unwanted calls it generates.

Of greatest interest to the call center and its clients, of course, is the sales rate.

In the matched groups test, the test group had 0.24 sales per representative per hour, versus 0.18 for the other group. This is about a one-third increase, which clients can easily be persuaded to find admirable.

In summary, Smart-Call breaks the traditional focus on dialer performance.

Instead, it concentrates, rightly, on improving the called parties’ experience, taking into account their demonstrated preferences about when they would prefer to be called. In fact, we are not at all convinced that the dialing system 2Contact used is all that good – but the tests and campaign experience using Smart-Call seem to indicate that the performance of the dialing system doesn’t matter all that much. Getting the timing right is much more important. (It should be noted that the numbers reported here have not been independently verified.)

If the success of this approach means that eventually more call centers will be calling you, and if that bothers you, we sympathize. As we said at the beginning, A NA L Y T I C S

the call and contact centers industry continues to thrive, although perhaps not growing as fast as it did earlier, because, despite all the complaints, these calls do work. On the other hand, if contact centers use this new technology, they seem quite likely to bother you less than they do now, even while making more calls.

They can achieve this by nearly eliminating nuisance calls in which you answer and they hang up because they have no one to talk to you, and by recognizing when you’re more likely to be receptive.

At least they will know not to keep calling you at dinnertime. ❙ Douglas A. Samuelson (samuelsondoug@yahoo. com), 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 contributing editor of OR/MS Today and Analytics magazines and a longtime member of INFORMS. Olcay Yucekaya (oyucekaya@2contact.com) is a data scientist at 2Contact who also​works with major companies outside the scope of 2Contact. He holds a master’s degree in marketing management from Tilburg University in Tilburg, the Netherlands. He gratefully acknowledges the assistance of Professor George Knox at Tilburg University in developing Smart-Call.

REFERENCES 1. Douglas A. Samuelson, “System for Regulating Arrivals of Customers to Servers,” U. S. Patent No. 4,858,120, Aug. 15, 1989.

2. Douglas A. Samuelson, “Predictive Dialing for Outbound Telephone Call Centers,” Interfaces, September-October 1999.

3. Michael Kaiser-Nyman, D. Samuelson and B. Swieskowski, “Predictive Dialing System Using Simulation,” U.S. Patent No. 9,088,650, July 21, 2015.

S E P T E M B E R / O C T O B E R 2 017

|

47

C OM PAN Y P RO F I LE

Air Liquide Multinational company’s Computational and Data Science R&D team supports an extensive, varied, dynamic research portfolio.

BY JEFFREY E. ARBOGAST (left) AND ATHANASIOS KONTOPOULOS

T

he world leader in gases, technologies and services for industry and health, Air Liquide is present in 80 countries with approximately 67,000 employees and serves more than 3 million customers and patients. Oxygen, nitrogen and hydrogen are among the essential small molecules for life, matter and energy. Essential small molecules embody Air Liquide’s scientific territory and have been at the core of the company’s activities since its creation in 1902.

With the 2016 acquisition of Airgas, Air Liquide has entered a new phase in its development and growth. By combining the

48

|

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

Airgas omnichannel approach of physical branch locations, e-business and telesales with Air Liquide innovation capabilities (including digital), the Air Liquide Group is accelerating its customer-centric transformation.

Air Liquide serves a wide range of industrial sectors, including automotive and manufacturing, food and pharmaceuticals, oil and gas, steelmaking, semiconductor fabrication and chemical production along with research and technology organizations, including universities. Air Liquide offer its customers – from independent craftsmen to large industrial companies – a range of solutions: industrial and specialty

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

Air Liquide’s Delaware Research and Technology Center in the United States.

Photo credit: Don Pearse Photographers, Inc.

gases, application technologies, welding and safety equipment and related services. Present in the continuum of healthcare from hospital to home, Air Liquide serves 1.4 million patients around the world.

The variety of essential small molecules that Air Liquide offers (including specialty gases, mixtures and new molecules that it designs and manufactures) and the diversity of its customer base requires various product supply chains. Air Liquide’s cylinder fill plants package gases for delivery to its customers and patients.

For higher volume and cryogenic needs, Air Liquide delivers liquefied gases in bulk to permanent on-site storage tanks at its customers’ locations. For its largest volume customers, Air Liquide directly supplies gas from its production plants via extensive pipeline networks in the world’s largest industrial basins (e.g., the U.S. Gulf Coast and the Benelux region in Europe).

A NA L Y T I C S

As articulated in its NEOS strategic program for 2016-2020, Air Liquide’s customer-centric transformation is founded upon four strategic pillars: operational excellence, selective investments, open innovation and a worldwide network organization. In Air Liquide’s Computational & Data Science R&D Global Lab, we apply our innovative and entrepreneurial spirit – the creative oxygen that sustains our growth – and advanced methods in operations research (O.R.) and analytics to support these strategic pillars and address our common challenges in the energy and environmental transition, the healthcare evolution and the digital transformation.

O.R. AND ANALYTICS AT AIR LIQUIDE Over the years, Air Liquide’s activities in O.R. and analytics have broadened from advanced process control to optimization

S E P T E M B E R / O C T O B E R 2 017

|

49

COM PAN Y P RO F I LE : A I R LI Q U IDE

and data science. As the focus evolves from operational excellence (e.g., efficiency and reliability) to support business growth (including sales, marketing and finance), the need for broadening expertise continues with a focus on statistics and financial economics. In research groups located in Paris, Delaware and Shanghai, we in Computational & Data Science R&D serve at the scientific core of these activities, engaging with a range of internal and external partners to remain at the forefront of technology. In addition to O.R. and analytics activities, our research includes connected devices, computational fluid dynamics and molecular modeling. Our research portfolio extends from exploratory research in line with broad scientific and technology trends impacting our business to key projects focused upon clear and present business and customer needs.

At Air Liquide, innovation in O.R. and analytics is a collaborative effort beyond R&D. Our Alizent entity, dedicated to Industrial IoT, develops, deploys and supports interactive monitoring and control solutions for returnable assets (e.g., cylinders), remote assets (e.g., customer storage tanks, delivery vehicles/drivers), and production assets of internal and external clients alike. Our digital fabs foster cross-disciplinary collaboration with our operations on strategic topics with a particular focus on data and our customercentric transformation.

Air Liquideâ&#x20AC;&#x2122;s worldwide network of modern R&D centers provides a diverse environment for innovation with various national, scientific and business line backgrounds present. Our team members in O.R. and analytics come from a variety of backgrounds including

Future of Air Liquideâ&#x20AC;&#x2122;s Paris-Saclay Research Center in France.

Photo credit: Michel Remon of Golem Images

50

|

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

chemical, industrial, mechanical and electrical engineering along with statistics and computer science. Some team members follow their project work into business and management roles outside of R&D, while others pursue a technical career path recognized as experts in our international expertise program. In R&D, we have valuable opportunities to work among and across different business lines and operational entities, with whom close proximity is essential to the successful application of O.R. and analytics to benefit our customers and business.

Open innovation is one of the ways that R&D accelerates innovation. In Computational & Data Science R&D, we are particularly engaged with the innovation ecosystem in North America and Europe, sponsoring research at leading academic institutions including Carnegie Mellon University and Ă&#x2030;cole Polytechnique, among others. Our Paris-Saclay Research Center is engaged within the Paris-Saclay innovation cluster that includes leading universities, engineering and business schools, research institutes, and industrial research and development centers. Our Delaware Research & Technology Center is located in close proximity to Philadelphia and the innovation clusters of the northeastern United States. Through our collaboration with Virginia Tech, Air Liquide is A NA L Y T I C S

a leader among the industrial members of the Center for Excellence in Logistics and Distribution (CELDi), a multi-university research consortium. Our external engagement enables us to gain valuable insights from research and applications in the chemical and process industries and beyond.

A particular example of open innovation is our sponsorship of the 2016 ROADEF/EURO Challenge on the inventory routing problem (IRP), a key optimization challenge for both industry and the scientific community. The open challenge is organized every two years by ROADEF and EURO, the French and European counterparts to INFORMS respectively, to solve an industrially-relevant research problem proposed by the sponsor.

WORLDWIDE APPLICATIONS OF O.R. AND ANALYTICS In the past, the bulk of Air Liquideâ&#x20AC;&#x2122;s liquefied gas customers managed their own inventories and called-in orders, operating in a customer-managed inventory (CMI) context. Many have now transferred this responsibility to Air Liquide where this vendor-managed inventory (VMI) relationship enables improved insight on the real-time customer demand rate and inventory level, generally transmitted automatically to our operations center using remote telemetry.

S E P T E M B E R / O C T O B E R 2 017

|

51

COM PAN Y P RO F I LE : A I R LI Q U IDE

To achieve operational efficiencies and customer centricity enabled by the VMI relationship, we develop and deploy advanced enterprise planning and scheduling tools in a mixed VMI/CMI context.

Beyond the routing of tractor-trailers to fulfill given orders, such an IRP solution requires effective, reasonable forecasts of customer demand to determine both the timing and the quantity of each customer delivery. The IRP problem offers interesting challenges for both scientific research and industrial practice.

Our sponsorship of the ROADEF/ EURO Challenge continues to influence the direction of scientific research by providing industrially relevant benchmark cases that include key practical challenges missing from previously published work on the IRP. The insights that we continue to gain complement our own extensive history on the topic, which includes applications of ant colony optimization, local search heuristics and column generation.

Our optimization work extends to the tactical and strategic level in the design of bulk and packaged gas supply chains.

This includes the sizing of distribution fleets, the location of sites (production, filling and transportation depots), the allocation of customers to those sites and the sizing of on-site customer storage.

Entities worldwide of various sizes have applied these planning tools to promote 52

|

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

operational excellence in the reliable, efficient supply of product to Air Liquide’s customers. Enterprise-wide optimization in the context of production-distribution coordination is of particular interest.

Air Liquide’s “Smart and Innovative Operations” (SIO) program integrates digital tools to analyze operating data from its production plants, exemplified at the new remote operation and optimization center in Lyon, France. Deployed globally, the SIO program focuses on automating and centralizing operations, optimizing the performance of each site and anticipating malfunctions. Consistent, fundamental model templates facilitate the deployment of SIO applications in optimization and data analytics. We have built a core optimization engine in AIMMS that includes model templates for unit operations (e.g., compression, distillation and liquefaction) for users to combine into a “digital twin” of their particular plant or pipeline network. We share knowledge in internal conferences on advanced process control and optimization.

With the SIO program, our operations have established new “operation center analyst” roles with responsibility to use optimization tools to define the mediumterm production plans and data analysis tools to analyze energy consumption relative to similar conditions in the past, as well as to detect the weak signals that precede a malfunction on critical W W W. I N F O R M S . O R G

equipment (predictive maintenance). In-depth process knowledge and experience in production positions are general qualities that an analyst brings to the position.

With this role, the community around O.R. and analytics continues to grow within Air Liquide. The community of analysts is responsible Air Liquide bulk storage tank and cryogenic trailer for delivery.

Photo credit: Air Liquide to drive the improvement of digital tools and identify future efficiency are the backbone of Air Liquide’s close projects. proximity with customers in the United In Computational & Data Science States, offering small and large busiR&D, we contribute to the development nesses and individual entrepreneurs a of in-house tools and the evaluation of “one-stop shop” for their cylinder gas and commercial-off-the-shelf tools, helping to services needs, as well as related safety transform “black box” solutions into “glass and welding hard goods. The branches box” solutions. We find that this is critiare fully integrated with and are complecal to promote sustained buy-in and use mented by digital and telesales activities among the end-users of such tools. While to improve customer reach and service. we remain responsible for evaluating, deIn addition to generating direct sales veloping and introducing new optimization to customers independently, telesales and data science methods, responsibility teams support branches and field sales for broader deployment transitions to our teams, and drive e-business traffic to the operational counterparts – particularly for website. The combination of this strong the specific needs of Air Liquide’s pipeline Airgas infrastructure with our scientific operations. background in O.R. and analytics offers exciting opportunities to deliver value for our stakeholders and customers.

EXCITING FUTURE WITH O.R. AND ANALYTICS With extensive worldwide operations The extensive geographic network of comprised of various supply chains and more than 1,100 Airgas retail branches a broad customer base having varied A NA L Y T I C S

S E P T E M B E R / O C T O B E R 2 017

|

53

COM PAN Y P RO F I LE : A I R LI Q U IDE

Computational & Data Science R&D researchers collaborating.

applications of essential small molecules, Air Liquide offers unique and varied opportunities for O.R. and analytics research due to the volume and variety of data, decision variables and constraints.

Considering the pace of production and distribution operations, the high velocity of data and decisions will continue to be a challenge and an opportunity (e.g., real-time optimization of pipeline and plant operations, application of real-time traffic and up-to-date telemetry readings to adapt distribution schedules).

The veracity of data to be analyzed and applied in decision-support tools will continue to be a challenge as we apply greater volumes of data from a variety of sources (internal and external from various systems). Assuring the veracity of our solutions in light of the uncertainty inherent as we consider forecasts well into the future remains a key challenge.

54

|

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

This particularly requires scientific expertise in O.R. for such techniques as discrete event simulation, adjustable robust optimization and stochastic optimization.

At Air Liquide, innovation is part of an open, user-centric ecosystem based on sciPhoto credit: Air Liquide ence, technologies, customer experience and the incubation of new activities. The ongoing customer-centric transformation strategy is modifying ways of working, consuming and communicating. O.R. and analytics are a keystone of this transformation in managing assets and interacting with customers, with Computational & Data Science R&D responsible for evaluating and developing new methods to be transferred for widespread application throughout Air Liquide. ❙ Jeffrey E. Arbogast (jeffrey.arbogast@airliquide. com), Ph.D., is an Air Liquide International Expert in Computational & Data Science R&D based in Delaware and is Air Liquide’s representative to the INFORMS Roundtable. He is a member of INFORMS. Athanasios Kontopoulos, Ph.D., is an Air Liquide International Senior Expert and Scientific Director of the Computational & Data Science R&D Global Lab based in Paris-Saclay. The authors thank Jean André and Bin Yu, Air Liquide International Experts based in Paris-Saclay and Delaware, respectively, for their feedback and contributions to this article.

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

• Network With Your Professional Peers and Those Who Share Your Interests • INFORMS Connect, Our Online Community, Helps You Network With Your Colleagues Quickly • INFORMS Communities and Meetings Provide Unsurpassed Networking Opportunities • INFORMS Certification for Analytics Professionals (CAP®) • Build Your Professional Profile With a Leadership Role in INFORMS • INFORMS Career Center Provides You With the Industry's Leading Job Board

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

B EYON D MO O R E ’ S LAW

Linking quantum computing, DNA and healthcare

BY TARAN VOLCKHAUSEN oore’s Law, which states that processing speeds will double every two years as we cram more and more silicon transistors onto chips, has been faltering since the early 2000s when the law started to run up against fundamental limitations presented by the laws of thermodynamics.

While the chip industry, with Intel leading the charge, has found ways to sidestep the limitations up until now, many are now saying that despite the industry’s best efforts, the stunning gains in processor speeds will not be seen again by the simple application of Moore’s Law. In fact, there is

M

56

|

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

evidence to show that we are reaching the plateau for the number of transistors that will fit on a single chip. Intel has even suggested silicon transistors can only keep getting smaller during the next five years.

As a result, Intel has resorted to other practices to improve processing speeds, such as adding multiple processing cores.

However, these new methods are just a temporary solution because computing programs can benefit from multi-processors systems up until a certain point.

WHERE DO WE GO FROM HERE? No doubt, the end of Moore’s Law will certainly present headaches in the

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

The “death” of Moore’s Law is stirring heightened interest in quantum computing and other “supercomputer” technology.

Photo Courtesy of 123rf.com | © Hannu Viitanen

immediate future for the technology sector. But is the death of Moore’s Law really all bad news? The fact the situation is stirring heightened interest in quantum computing and other “supercomputer” technology gives us reason to suggest otherwise.

Quantum computers, for instance, do not rely on traditional bit processors to operate. Instead, quantum computers make use of quantum bits, known as “qubits,” a two-state quantum-mechanical system that can process both 1s and 0s at the same time.

The advances in processing speeds made possible by quantum computing would make Moore’s Law look like a caveman’s stone tool. For instance, the Google-funded D-Wave quantum supercomputer can outperform traditional

A NA L Y T I C S

computers in processing speeds by a mind-blowing factor of 100 million. With the advantages offered by “quantum supremacy” easy to comprehend, the race is now on between tech-heavyweights such as Google, IBM, Microsoft and Intel to successfully prototype and release the first quantum computer for commercial use.

However, due to the “weird” quantum mechanics the technology relies on, there are barriers to working with and storing data derived from processing with qubits.

QUANTUM COMPUTING WITH DNABASED STORAGE Basically, the fundamentals of quantum mechanics don’t permit you to store information on the quantum-computing

S E P T E M B E R / O C T O B E R 2 017

|

57

QUAN TU M C OM P U T I N G & D NA STORAGE

machine itself. While you could convert its data for storage on traditional devices, such as the solid-state hard drive, you would need to process a nearly infinite amount of information, which would require an impossible amount of space and energy to achieve.

However, there could be a solution, but it requires us to look within. Not in a hippy-dippy “finding yourself” sort of way, but rather the double helix code found in humans and almost all other organisms: DNA. For decades, researchers have toyed around with using DNA as both a computing and a storage device. Recently, a team of researchers at Columbia University demonstrated that their coding strategy based on one strand of DNA could store 215 petabytes of information.

“Performing sentiment analysis on quantum computing and DNA storage topics with Vector API, may uncover robust

demand for these technologies in various industries such as healthcare.” says Jo Fletcher, co-founder of Indexer.me.

SUPER COMPUTERS AND HEALTH TREATMENTS The human body is an incredibly complex organism. While the markets have released many life-saving drugs, many barriers hold us back from realizing their maximum potential.

Standard computing isn’t powerful enough to truly predict the ways a drug will react with an individual’s particular genetic composition and unique environmental factors. With quantum computing based on DNA storage, however, you would have the ability to examine pretty much any scenario imaginable by mapping a much more accurate prediction of any given drug’s interaction with a particular person based on their genetics and environment.

With quantum computing, medical professionals can open a new chapter in drug prescription outcomes by tailoring each treatment to meet the exact requirements of each individual. ❙

For decades, researchers have toyed around with using DNA as both a computing and a storage device.

Photo Courtesy of 123rf.com | © Anton Riakhin

58

|

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

Taran Volckhausen is a contributing editor at Vector, a natural language processing application that performs information extraction on millions of news stories per day.

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

2017 INFORMS ANNUAL MEETING

Our thoughts and prayers are with all those affected by Hurricane Harvey and the unprecedented rain and ďŹ&#x201A;ooding impacting Houston.

At this time, we are closely monitoring the situation in Houston to determine its impact on the INFORMS 2017 Annual Meeting currently scheduled there for October 22â&#x20AC;&#x201C;25.

We will provide updates to speakers and attendees as quickly as possible to ensure our guests can plan accordingly.

The Red Cross has launched a massive relief response to this devastating storm and needs financial donations to be able to provide immediate disaster relief. Help people affected by Hurricane Harvey by visiting redcross.org, calling 1-800-RED CROSS or texting the word HARVEY to 90999 to make a $10 donation.

http://meetings.informs.org/nashville2016 http://meetings.informs.org/houston2017

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

INFORMS 2017 Annual Meeting Conference will offer something for every analytics and O.R. practitioner and researcher.

EDITOR’S NOTE: Along with sending its thoughts and prayers, INFORMS salutes the people of Houston and all of those throughout Texas and beyond who have been responding in such an inspiring manner to the aftermath of the devastating hurricane and flooding. INFORMS is closely monitoring the situation in Houston to determine its impact on the 2017 Annual Meeting, and will provide updates as quickly as possible.

60

|

The 2017 INFORMS Annual Meeting is currently scheduled for Houston, Texas, on Oct. 22-25.

Houston is world renowned for its economy based in aeronautics, energy, manufacturing and transportation. Houston’s energy industry is recognized worldwide for its renewable energy sources, including wind and solar power. Houston is the most diverse city in Texas and has a large and growing international community. Because it is the home of the NASA Lyndon B. Johnson Space Center and its strong ties to the aeronautic industry, it has earned the nickname “Space City.” With a little of something for everyone, Houston lends itself as the perfect location for the next INFORMS Annual Meeting.

The meeting will take place at the George R. Brown Convention Center and the Hilton Americas, with all technical sessions taking place at the Convention Center. INFORMS also has group rates at the Marriott Marquis Houston and the DoubleTree by

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

Hilton Houston Downtown.

Only a limited number of rooms are blocked and they will sell out quick, so please make your reservations as soon as possible.

PRE-CONFERENCE WORKSHOPS This year INFORMS is introducing a new preconference workshop in conjunction with the Annual Meeting. The Academic Leadership Workshop will take place on Oct. 21, the day before the start of the Annual Meeting. The allday event is designed for faculty of all ranks with an interest in every level of academic leadership. The workshop provides information that can be useful for becoming efficient and As always, the 2017 INFORMS Annual Meeting will draw a large and diverse group of effective leaders. Panel attendees from throughout the world.

Image Courtesy of INFORMS speakers are highly visible world-class current and former acaAnother pre-conference workshop demic administrators. Participants will that is being introduced this year is the have an opportunity to ask questions INFORMS Workshop on Data Science. and network with peers in academic Sponsored by the INFORMS College on leadership positions. All participants Artificial Intelligence, this workshop is must be nominated by a department a premier research event dedicated to head/chair or college dean. developing novel data science theories,

A NA L Y T I C S

S E P T E M B E R / O C T O B E R 2 017

|

61

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

algorithms and methods to solve challenging and practical problems that benefit business and society at large. The conference invites innovative data science research contributions that address business and societal challenges from the lens of statistical learning, data mining, machine learning and artificial intelligence. Contributions on novel methods may be motivated by insightful observations on the shortcomings of state-of-theart data science methods in addressing practical challenges, or may propose

entirely novel data science problems.

Research contributions on theoretical and methodological foundations of data science, such as optimization for machine learning and new algorithms for data mining, are also welcome.

NETWORKING Along with the abundance of educational opportunities, the conference will offer several opportunities for connecting and networking. The Welcome Reception will be held on Oct. 22. Subdivision

WSC turns 50: simulation everywhere!

BY ERNEST H. PAGE The Winter Simulation Conference (WSC) was first held in 1967 as the Conference on the Applications of Simulation Using GPSS. As we celebrate its 50th anniversary this year, we are proud that the conference is widely regarded today as the premier international forum for disseminating recent advances in the field of dynamic systems modeling and simulation. In addition to a broadly scoped, high-quality technical program, WSC is the annual meeting place for simulation researchers, practitioners and vendors spanning a wide array of disciplines and working across industry, government, military, service and academic sectors. This year’s conference will be held Dec. 3-6 in Las Vegas at the Red Rock Resort.

A hallmark of WSC is the diversity of its program, and this year’s program chair, Gabriel Wainer, has organized

62

|

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

a fantastic collection of tracks. In recognition of the 50th anniversary milestone, we DECEMBER 3–6, 2017 will have a formidable “History of Simulation” track. But we will also be intently looking to the next 50 years in our “Future of Simulation” track. Of course, many of the traditional tracks will return for 2017, including academically focused tracks: Analysis Methodology, Modeling Methodology, Simulation Optimization, AgentBased Simulation, Education and Hybrid Simulation, as well as applied tracks such as Aviation Modeling and Simulation, Healthcare Applications, Logistics, Supply

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

meetings will be held predominantly on Oct. 23 in the evening. On Oct. 24, INFORMS will host the General Reception at Minute Maid Park, Home of the Houston Astros.

Another unique networking opportunity for student members is the Coffee with a Member program. This wonderful program connects INFORMS students with some of INFORMS most enthusiastic members for 15-minute impromptu meetings and some sage INFORMS advice. We know the Annual

Chain Management and Transportation, Manufacturing Applications, Modeling and Analysis of Semiconductor Manufacturing (MASM), Military Applications and Homeland Security, Project Management, Construction, Cyber Physical Systems and Intelligent, Adaptive and Autonomous Systems. The program also offers a great mix of introductory and advanced tutorials, case studies, a Ph.D. colloquium, poster sessions and vendor presentations. Whatever your interests, the WSC 2017 program will have something for you.

We have a great lineup of keynote speakers for WSC 2017. Barry Nelson, Northwestern University, will set the stage for the conference by asking the question “WSC 2067 - What Are the Chances?” in the opening keynote.

Our military keynote speaker is Douglas Hodson, Air Force Institute of Technology, and our MASM keynote will be given by Stéphane Dauzère-Pérès, École Nationale Supérieure des Mines de Saint-Étienne. The 2017 “Titans of Simulation,” a distinguished speaker lunchtime series, will feature professors Robert G. Sargent and Bernard P. Zeigler.

The Winter Simulation Conference is jointly sponsored by the American Statistical Association (ASA), Arbeitsgemeinschaft Simulation (ASIM), Association

A NA L Y T I C S

Meeting can be a bit overwhelming and hope these casual meetings will make students more comfortable, knowledgeable and enthusiastic about both the meeting and INFORMS. Space is limited and open to first-time attendees/ participants only. Students may enroll for this program when they register for the meeting.

CAREER CENTER A huge benefit of the INFORMS Annual meeting for employers and job

for Computing Machinery: Special Interest Group on Simulation (ACM/SIGSIM), Institute of Electrical and Electronics Engineers: Systems, Man, and Cybernetics Society (IEEE/SMC), Institute for Operations Research and the Management Sciences: Simulation Society (INFORMS-SIM), Institute of Industrial and Systems Engineers (IISE), National Institute of Standards and Technology (NIST), and the Society for Modeling and Simulation International (SCS), and operates under the direction of the conference Board of Directors, whose volunteer members represent these sponsoring organizations and ensure the continuity, quality and traditions of operation that make WSC a truly unique conference.

From the program to the extensive collection of exhibitors and vendors, to the meetings of many professional societies and user groups, to the various social gatherings, to the beautiful Red Rock Resort, and the myriad extracurricular options in the Las Vegas area, WSC 2017 promises to be a fantastic event. We hope you will plan on joining us!

For more information, visit http://wintersim.org. ❙

Ernest H. Page is general chair of WSC 2017.

S E P T E M B E R / O C T O B E R 2 017

|

63

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

seekers is the INFORMS Career Center.

The Career Center and activities provide employers with the opportunity to meet and collect resumes from numerous job seekers in a short period of time, early in the meeting, and to schedule and set-up private interviews later in the meeting.

Career Center activities are free for all individual meeting registrants.

Job seekers should register in the INFORMS Career Center so employers will know you are attending. Be sure to post your resume or anonymous career profile that will lead employers to

you. Job seekers should review Career Center Resources prior to attending the job fair to make sure your resume and interviewing skills are in tip-top shape.

We hope you will attend this unique opportunity to connect and network with more than 5,000 INFORMS members, students, prospective employees, and academic and industry experts. We look forward to seeing you in Houston! ❙ For more information on any of the events listed in this article or other activities at the Annual Meeting, visit www.meetings.informs.org/houston2017.

CALL FOR APPLICATIONS 2018

DEADLINE FOR APPLICATIONS IS TUESDAY, OCTOBER 31, 2017 Are you the best in OR/MS/Analytics education?

The UPS George D. Smith Prize is created in the spirit of strengthening ties between industry and the schools of higher education that graduate young practitioners of operations research. INFORMS, with the help of The Practice Section, will award the prize to an academic department or program for the effective and innovative preparation of students to be good practitioners of operations research, management science, or analytics.

The prize will include a trophy and $10,000 award. The UPS George D. Smith Prize will be announced at the 2018 Edelman Gala at the INFORMS Conference on Business Analytics and Operations Research, in Baltimore, MD.

For more information, questions can be sent to Rina Schneur, 2018 Smith Prize Chair at rinarsg@gmail.com.

www.informs.org/smithprize 64

|

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

2017 UPS Smith Prize Winners United States Air Force Academy, Operations Research Program

g ties enin h t eng str

ups george d smith prize W W W. I N F O R M S . O R G

SPE CIAL ADVE RT ISIN G SE C T IO N

FACULTY POSITIONS #051186 AND #49923 IN INFORMATION TECHNOLOGY AND OPERATIONS MANAGEMENT

C L AS S IF I E D S

FACULTY POSITIONS #051186 AND #49923 IN INFORMATION TECHNOLOGY AND OPERATIONS MANAGEMENT

SMU’s Cox School of Business invites applications for full-time faculty positions at the level of Assistant Professor in Information Technology and Operations Management (ITOM). The ideal candidates would possess strong quantitative/analytical skills and a Ph.D. in information systems or a related field. The ITOM Department offers courses in the School’s BBA, MBA, MS and EMBA programs. The positions begin in Fall 2018.

SMU’s Cox School of Business invites applications for a full-time Clinical faculty position, at the level of Clinical Professor in Information Technology and Operations Management (ITOM). The ideal candidate would possess strong quantitative/analytical skills and a Ph.D. in information systems or a related field. The ITOM Department offers courses in the School’s BBA, MBA, MS and EMBA programs. The position begins Fall 2018.

The Cox School is a nationally ranked business school located in Dallas, Texas, the premier business center in the US Southwest. The ITOM department has a well-respected research faculty and excellent relations with the corporate and business community, providing a unique and exciting environment for high-quality research. The School offers a collegial working environment, generous faculty support and outstanding facilities. The Dallas Fort-Worth (DFW) Metroplex offers a thriving business community in one of the fastest-growing regions of the country, a relatively low cost of living and myriad cultural and recreational activities and resources. Information about the School can be found at http://www.smu.edu/cox.

The Cox School is a nationally ranked business school located in Dallas, Texas, the premier business center in the US Southwest. The ITOM department has a well-respected research faculty and excellent relations with the corporate and business community, providing a unique and exciting environment for high-quality research. The School offers a collegial working environment, generous faculty support and outstanding facilities. The Dallas Fort-Worth (DFW) Metroplex offers a thriving business community in one of the fastest-growing regions of the country, a relatively low cost of living and myriad cultural and recreational activities and resources. Information about the School can be found at http://www.smu.edu/cox.

Priority will be given to applications received by October 1, 2017, although the search committee will continue to accept applications until the positions are filled. The ITOM Faculty will be conducting initial interviews at the INFORMS National Meeting and the Conference on Information Systems and Technology in Houston, Texas in October 2017.

Applications must be submitted electronically to http://apply.interfolio.com/43369

Priority will be given to applications received by October 1, 2017, although the search committee will continue to accept applications until the position is filled. The ITOM Faculty will be conducting initial interviews at the INFORMS National Meeting and the Conference on Information Systems and Technology in Houston, Texas in October 2017.

Applications must be submitted electronically to https://apply.interfolio.com/43370

SMU will not discriminate in any program or activity on the basis of race, color, religion, national origin, sex, age, disability, genetic information, veteran status, sexual orientation, or gender identity and expression. The Executive Director for Access and Equity/Title IX Coordinator is designated to handle inquiries regarding nondiscrimination policies and may be reached at the Perkins Administration Building, Room 204, 6425 Boaz Lane, Dallas, TX 75205, 214-768-3601, accessequity@smu.edu.

SMU will not discriminate in any program or activity on the basis of race, color, religion, national origin, sex, age, disability, genetic information, veteran status, sexual orientation, or gender identity and expression. The Executive Director for Access and Equity/Title IX Coordinator is designated to handle inquiries regarding nondiscrimination policies and may be reached at the Perkins Administration Building, Room 204, 6425 Boaz Lane, Dallas, TX 75205, 214-768-3601, accessequity@smu.edu.

SMU is an Affirmative Action/Equal Opportunity Institution

SMU is an Affirmative Action/Equal Opportunity Institution

informs career fair The INFORMS CAREER CENTER offers employers expanded opportunities to connect to qualified O.R. & analytics professionals. INFORMS offers a complete line of services to be used alone or in conjunction with the Career Fair at the 2017 Annual Meeting, giving job seekers and employers a convenient venue to connect. The Career Fair is free to INFORMS attendees.

EMPLOYERS PARTICIPATING IN THE CAREER FAIR ACTIVITIES AT THE 2017 ANNUAL MEETING WILL BE ABLE TO: • Provide their recruitment materials in a fun and energetic Career Fair setting • Schedule their own on-site interviews at reserved tables or interview booths • Promote their organization and meet highly-qualified, diverse candidates FOR CAREER FAIR REGISTRATION & INFORMATION:

http://meetings.informs.org/houston2017

A NA L Y T I C S

S E P T E M B E R / O C T O B E R 2 017

|

65

FIVE- M IN U T E A N A LYST

Nuclear deterrence The study of deterrence is difficult because the data is censored in the sense that you only receive data when deterrence fails.

BY HARRISON SCHRAMM

66

|

I have a longstanding policy of not commenting on current events in this column, but rather keeping it “light,” with use cases involving everyday problems, popular television shows and, for quite a bit of 2016, Star Wars. Originally, I sat down to write this column about “a world without the central limit theorem,” which was going to have this tagline: Tonight, we’ve replaced Bob’s Normal Distribution [1] with the Cauchy – let’s see what happens. Bob just had 8,000 cups of coffee – he never does that!

That was the article that I was going to write, and probably will for November. However, the current escalation between the United States and North Korea [2] caused me to put that on “pause” and return to a topic that I used to think about quite a bit previously – deterrence. There isn’t enough room to do this topic justice in one column, but I’d like to (repeat) some ideas that may be worth thinking about.

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

The study of deterrence is difficult because the data is censored in the sense that you only receive data when deterrence fails. One could – conceivably – count every day that nations did not go to war as a “success.” Conversely, one could claim times that the world was at the “brink” – such as the Cuban Missile Crisis – to be deterrence “successes”; however, it becomes exceedingly difficult to know when these brinks truly happen.

After all, in the nuclear age, a politician late at night could be every bit as dangerous as any overt military buildup.

There are several excellent books written about deterrence, all worth a read, particularly “Arms and Influence”

by Thomas Schelling and “On Thermonuclear War” by Hermann Kahn. Of course, the classic film “Dr. Strangelove,” which was inspired in part by the aforementioned Kahn, is always worth watching. In this column, I’d like to point out one very simple “nuclear war” game. In it, we’re going to take a continuous decision process and make it a discrete (finite valued) one. For ease of analysis, each side will register their decision at a fixed time,

say 8 a.m., and their payoff will be immediately adjudicated.

The choices are “launch,” which means initiate an attack, or “wait,” in which case a player does not launch their weapon but will retaliate if launched against (Table 1).

I haven’t filled in the boxes yet, but will explain as follows: First, the “null” case. If both sides play “wait,” then nothing happens, and the world goes on until the next morning.

Second, the “launch/launch” case. If both sides launch, then they will both incur the damage the other will inflict. For sake of simplicity, let’s make this: < -10,-10 >.

Now, the interesting case. What happens if one side plays “launch” and the other plays “wait”? It might seem at first that the player who launches has the upper hand, because he will be able to inflict damage on the other side first and possibly eliminate the other player’s weapons – a so-called “splendid first strike.”

Developed powers are protected from this sort of strike by remote/airborne bombers and submarines.

Table 1: One-shot deterrence game with unknown payoffs.

Player 1 / Player 2

Launch

Wait

Launch

<,>

<,>

Wait

<,>

<,>

A NA L Y T I C S

S E P T E M B E R / O C T O B E R 2 017

|

67

FIVE- M IN U T E A N A LYST Table 2: Payoff table for the one-shot deterrence “game.”

Player 1 / Player 2

Launch

Wait

Launch

< -10,-10 >

< 10,-10 >

Wait

< -10,10 >

< 0,0 >

Table 3: The asymmetric nuclear deterrence game.

Player 1 / Player 2

Launch

Wait

Launch

< 10,-10 >

< 10,-10 >

Wait

< -10,10 >

< 0,0 >

The red arrows in Table 2 represent the fact that with second-strike options, the one-sided “launch” choices are indistinguishable from the “launch, launch” cell. Neither side can unilaterally do better than the <0,0> payoff. And the world goes on.

If one side has a “small” number of nuclear weapons, they may perceive that the other side has the advantage.

In Table 3, Player 2 perceives – correctly – that Player 1 can always do better by choosing “launch” over “wait.”

Given that Player 1’s optimum choice is to “launch,” Player 2 may choose to use their own weapons.

It is for this reason that – paradoxically – having a more diversified portfolio of nuclear weapons makes one less likely

to feel the need to use them. If this five-minute piece leaves you with one idea, it is that the notion of “use/lose” is dangerous and should be avoided.

A ‘final’ thought: It is estimated that the time of flight of an ICBM is around 40 minutes on the high-side. If the missiles were launched when you started reading, you would have around 35 minutes left. ❙ Harrison Schramm (Harrison.schramm@gmail. com), CAP, PStat, is a principal operations research analyst at CANA Advisors, LLC, and a member of INFORMS.

REFERENCES 1. This references a series of 1980s advertisements for Folger’s Crystals.

2. https://www.washingtonpost.com/news/ worldviews/wp/2017/08/08/trumps-fire-and-furystatement-echoes-north-koreas-own-threats

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

68

|

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

Disney Working at is unlike anything else. That’s because we’re always finding new ways to interact with fans – wherever and however they connect with sports. When you have the latest technology, game-changing ideas and world-class talent on your team, every day is extraordinary. ESPN is setting the standard for data as it relates to sports fans.

ESPN’S GLOBAL DATA MANAGEMENT TEAM IS: PASSIONATE

COLLABORATIVE

INNOVATIVE

CREATIVE

Meet Christy | Director of Customer Relationship Management at ESPN AS A DATA PROFESSIONAL, WHAT ATTRACTED YOU TO WORK AT ESPN?

Sports and data really do go hand in hand. ESPN has a wealth of historical data on sports, athletes, games, injuries – you name it. It’s a place full of data.

What I saw was an opportunity to capitalize on the fan relationship and ESPN’s one-to-one relationship with fans based on their favorite teams, the sports that they care about, and their personal passions and interests.

HOW DOES YOUR TEAM CONTRIBUTE TO THE LARGER ESPN STORY?

My team is special. We really do start with the fan relationship. Our goal is to know enough about sports fans to serve them anytime and anywhere.

It’s really about serving them better. We have done a lot of work over the years to figure out how to tailor our marketing so we are hitting the right fans with the right message. Whether we’re helping them tune into their favorite game on WatchESPN or encouraging them to play fantasy football, every message is driven by data. We’ve made a difference for the business by analyzing our relationship with our fans.

BE A PART OF OUR STORY

Disney Careers

DisneyDataJobs.com/Informs

Disney Careers

EOE • DRAWING CREATIVITY FROM DIVERSITY • ©DISNEY

The Walt Disney Company

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

Unlock the decoded message DKFFQHXAQBUTKTKUGHJNMDHPAQBDUCKO TUOEKGMDKOQGKVHMDAQBTPBSKTHQTUJU FAMHOPPEHFFPHVUJMKGMQMDUJEAQBXQTT KUGHJNMDKSBYYFQTXQTMDKPKSUPMMKJA KUTPUJGFKMAQBEJQVMDUMMDHPVHFFRKM DKFUPMSBYYFKMDHPDUPRKKJUVQJGKTXBF QSSQTMBJHMAXQTIKMQOQJMTHRBMKMQMD KHJXQTIPOQIIBJHMAUJGIKKMIUJAJKVSKQS FKRBMHMPMHIKMQOFQPKMDHPODUSMKTUJ GIQCKQJMQJKVQSSQTMBJHMHKPMDUJEAQB

The final column: What does it say?

The accompanying message is encrypted with a simple substitution cypher (one letter replaced with another). To make this puzzle more challenging, all punctuation has been removed.

As a hint, the following words are in the unencrypted message: PUZZLOR, ANALYTICS and INFORMS. QUESTION: What does the unencrypted message say?

Send your answer to puzzlor@gmail.com by Nov. 15. 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

70

|

John Toczek is the senior manager of analytics at NRG in Philadelphia. He earned his BSc. in Chemical Engineering at Drexel University (1996) and his MSc. In operations research from Virginia Commonwealth University (2005).

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

O B J E C T - O R I E N T E D A P P L I C AT I O N PROGRAMMING INTERFACES The object-oriented GAMS APIs allow the smooth integration of GAMS into other programming environments by providing appropriate classes for the interaction with GAMS. The GAMSDatabase class provides in-memory representation of data for convenient exchange of input data and model results.

The GAMSJob class executes GAMSmodels.

The GAMSOptions class customizes GAMS models through the API. The GAMSModelinstance class solves sequences of closely related model instances in the most efficient way.

Available for .NET, Java, Python, and (new) C++ (open source) For more information, technical documentation, and examples see www.gams.com/latest/docs/apis

WHAT CUSTOMERS THINK ABOUT THE APIs

“

GAMS API is a very good way to encapsulate GAMS models inside programming languages. This API allows to have a dynamic link between the GAMS model and our applications. It is very robust and efficient. Moreover it includes a new feature that allows to solve several close instances of a same problem very fast. With this feature, we implemented very efficient sensitivity analysis of our models.

Dimitri Tomanos, Modeller analyst, GDF-Suez

”

“

With the GAMS .NET API we were able to implement some complex recursive MIP-based algorithms we could not easily express in the GAMS language itself. One advantage of the GAMS API was that we could reuse large parts of database access and data manipulation steps implemented in GAMS. Erwin Kalvelagen, Amsterdam

”

Optimization Modeling Group

High-level algebraic modeling system which incorporates all major commercial and academic state-of-the-art solvers

sales@gams.com www.gams.com Phone: (202) 342 0180