HTTP://WWW.ANALYTICS-MAGAZINE.ORG JANUARY/FEBRUARY 2015 DRIVING BETTER BUSINESS DECISIONS BROUGHT TO YOU BY: DATA VISUALIZATION • What’s next: Seeing is believing • Realization: Big picture of big data Executive Edge Bill Gossman, head of ASA Corp., on why leadership is the real game- changer for using data analytics ALSO INSIDE: • 2015: Predictions for predictive analytics • Small analytics company takes on Ebola • Case study: Why cloud analytics works
Transcript
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
januaRy/febRuaRy 2015DRiving betteR business DecisiOnsbROught tO yOu by:
Data Visualization• What’s next: seeing is believing• Realization: Big picture of big data
Executive Edge
bill gossman, head of asa corp., on why leadership is the real game-changer for using data analytics
also insiDE:• 2015: Predictions for predictive analytics• small analytics company takes on Ebola• Case study: Why cloud analytics works
w w w. i n f o r m s . o r g2 | a n a ly t i c s - m aga z i n e . o r g
Prediction: Big year for analytics
insiDe stORy
Happy New Year! ’Tis the season to be making predictions about what the future holds for the analytics profession, and who better to do that than a trio of experts in pre-dictive analytics who contributed their in-sights to this issue of Analytics magazine?
Glenn Wegryn, president of the Analytics Section of INFORMS and prin-cipal at Analytic Impact LLC, is the first to look into the crystal ball for his column, “Top 5 analytics predictions for 2015.” Wegryn sees another winning year for the profession as the demand for analyt-ics skills continues to grow. Wegryn ex-pects analytics programs to continue to pop up at business schools throughout the United States in response to the de-mand, and that the INFORMS Certified Analytics Professional program will in-crease in importance as a certification of skills and experience.
Wegryn also expects the analytics community to move closer to a common definition of “analytics,” bridging the gap between the data-centric definition (What can the data tell us?) and the decision-centric version (What is the problem we’re trying to solve?).
Next, Andrew Jennings, chief analytics officer at FICO, offers “Five more predic-tions for 2015.” Jennings starts by noting that since the hype around big data has
finally crested, big data and analytics will become “business as usual” and “more companies will cultivate in-house analytic competencies and many will embrace an-alytics as an important part of their com-pany cultures.”
Our third prognosticator, independent consultant and business advisor Rajib Ghosh, focuses on the healthcare in-dustry in his column, “What to expect in 2015.” Ghosh predicts a “great year for healthcare analytics professionals,” and that providers, not payers, will drive the market for analytics.
“Gaining market share has become the new mantra for the insurance compa-nies,” Ghosh writes. “They are also busy trying to find new business models for the future. Big providers are under immense pressure to improve their bottom line in the new world of the Affordable Care Act and government-driven payment reform.”
As for me, I’ll pass on making any predictions about what the future holds for analytics, both the profession and the online magazine you’re now reading. Instead, I’ll quote another scientist you might have heard of:
“I never think of the future, it comes soon enough.” – Albert Einstein
w w w. i n f o r m s . o r g4 | a n a ly t i c s - m aga z i n e . o r g
DRIVING BETTER BUSINESS DECISIONS
c o n t e n t s
FEatuREsleaDeRship: the Real game changeRleaders who take a proactive initiative to data analytics set the stage for successful, actionable use of information.By Bill gossman
the futuRe Of Data visualizatiOnnextgen technologies and evolving cognitive frameworks move data visualization from art to science.By will towler
Data visualizatiOn: big pictuRe Of big Datarealizing how quickly we understand what we see makes data visualization a key component of better decision-making.By nana s. Banerjee
analytics helps cOntain ebOlaHow one small data science and analytics company found a way to help contain the spread of the deadly virus.By Douglas a. samuelson and Brian Umana
ANALYTICS IN YOUR BROWSEROptimization and Simulation in Excel Online and Google Sheets
Advanced Analytics built on Business Intelligence is moving to the Web. But you don’t have to learn a complex new language to create your analytic models. Start and finish today in Excel Online or Google Sheets.
Use our Analytics Apps in your Web Browser
Solve all types of optimization models with Frontline’s free Solver App, for Excel Solver features in Excel Online and Google Sheets. To solve large-scale optimization models, upgrade to our Premium Solver App.
Run Monte Carlo simulation models with Frontline’s Risk Solver App – free for limited-size models, and free without size limits for desktop Risk Solver Pro, Risk Solver Platform, and Analytic Solver Platform users.
Use our free XLMiner Data Visualization App to explore your data in Excel Online – including Pivot Table results drawn from the powerful multidimensional data model in Office 365 Power BI and desktop Excel 2013.
Powerful Analytics for Your Laptop or Desktop
You can create your models in your web browser, but there’s nothing like a big screen and a fast CPU plus software to develop sophisticated analytic models. And we make it easy to publish your analytic models from desktop Excel to either Excel Online or Google Sheets.
Use Premium Solver Platform for industrial-strength conventional optimization – from linear, quadratic and mixed-integer programming to nonlinear, non-smooth and global optimization.
Use Risk Solver Platform for super-fast Monte Carlo simulation, decision trees, simulation optimization, robust optimization and stochastic programming.
Use XLMiner Platform for forecasting and data mining, from time series methods to classification and regression trees, neural networks and association rules.
Or use Analytic Solver Platform to do it all in Excel – building on Microsoft’s Power Pivot, Power Query, and Power View add-ins to work with 100 million row data sets from enterprise and cloud databases.
Find Out More, Download Your Free Trial Now
Visit www.solver.com/apps to learn more, register and download a free trial – or email or call us today.
The Leader in Analytics for Spreadsheets and the WebTel 775 831 0300 • [email protected] • www.solver.com
6 |
DRIVING BETTER BUSINESS DECISIONS
RegisteR foR a fRee subscRiption:http://analytics.informs.org
infoRMs boaRd of diRectoRs
President L. Robin Keller, University of California, Irvine
President-Elect Edward H. Kaplan, Yale University Past President Stephen M. Robinson, University of
Wisconsin-Madison Secretary Brian Denton, University of Michigan Treasurer Sheldon N. Jacobson, University of Illinois Vice President-Meetings Ronald G. Askin, Arizona State University Vice President-Publications Jonathan F. Bard, University of Texas at Austin Vice President- Sections and Societies Esma Gel, Arizona State University Vice President- Information Technology Marco Lübbecke, RWTH Aachen University Vice President-Practice Activities Jonathan Owen, CAP, General Motors Vice President-International Activities Grace Lin, Institute for Information Industry Vice President-Membership and Professional Recognition Ozlem Ergun, Georgia Tech Vice President-Education Jill Hardin Wilson, Northwestern University Vice President-Marketing, Communications and Outreach E. Andrew “Andy” Boyd, University of Houston Vice President-Chapters/Fora David Hunt, Oliver Wyman
infoRMs offices
www.informs.org • Tel: 1-800-4INFORMS Executive Director Melissa Moore Meetings Director Laura Payne Communications Director Barry List Headquarters INFORMS (Maryland) 5521 Research Park Drive, Suite 200 Catonsville, MD 21228 Tel.: 443.757.3500 E-mail: [email protected]
analytics editoRial and adveRtising
Lionheart Publishing Inc., 506 Roswell Street, Suite 220, Marietta, GA 30060 USATel.: 770.431.0867 • Fax: 770.432.6969
President & Advertising Sales John Llewellyn [email protected] Tel.: 770.431.0867, ext. 209 Editor Peter R. Horner [email protected] Tel.: 770.587.3172 Assistant Editor Donna Brooks [email protected] Art Director Jim McDonald [email protected] Tel.: 770.431.0867, ext. 223 Advertising Sales Sharon Baker [email protected] Tel.: 813.852.9942
AnAlytic Solver PlAtformeasy to Use, industrial Strength Predictive Analytics in excel
How can you get results quickly for business decisions, without a huge budget for “enterprise analytics” software, and months of learning time? Here’s how: Analytic Solver Platform does it all in Microsoft Excel, accessing data from PowerPivot and SQL databases.
Sophisticated Data Mining and Predictive Analytics
Go far beyond other statistics and forecasting add-ins for Excel. Use classical multiple regression, exponential smoothing, and ARIMA models, but go further with regression trees, k-nearest neighbors, and neural networks for prediction, discriminant analysis, logistic regression, k-nearest neighbors, classification trees, naïve Bayes and neural nets for classification, and association rules for affinity (“market basket”) analysis. Use principal components, k-means clustering, and hierarchical clustering to simplify and cluster your data.
Simulation, Optimization and Prescriptive Analytics
Analytic Solver Platform also includes decision trees, Monte Carlo simulation, and powerful conventional and stochastic optimization for prescriptive analytics.
Help and Support to Get You Started
Analytic Solver Platform can help you learn while getting results in business analytics, with its Guided Mode and Constraint Wizard for optimization, and Distribution Wizard for simulation. You’ll benefit from User Guides, Help, 30 datasets, 90 sample models, and new textbooks supporting Analytic Solver Platform.
Surprising Performance on Large Datasets
Excel’s ease of use won’t limit what you can do – Analytic Solver Platform’s fast, accurate algorithms rival the best-known statistical software packages.
Find Out More, Download Your Free Trial Now
Visit www.solver.com to learn more, register and download a free trial – or email or call us today.
w w w. i n f o r m s . o r g8 | a n a ly t i c s - m aga z i n e . o r g
item: I recently spent an afternoon visiting the Bay Area Advanced Manufacturing hub. Located in a long-shuttered auto manufacturing plant, BAAM brings to-gether a dozen or so companies that are all engaged in activities that involve 3D printing, a world that has been growing rapidly for the last several years, albe-it from a small base. Filling a large industrial space that has been largely empty for decades (the bottom floor was converted into a strip mall sometime in the 1980s), the BAAM hub features a variety of firms, in-cluding developers of 3D scanners, 3D design firms, companies specializing in material procurement and recycling, a 3D design software training firm, and a full service production house for complex physical items.
The lead company in this hub is Type A Machines, a company whose specialty is the development and deployment of 3D printers. Type A Machines’ CEO Es-pen Sivertsen gave me a very interesting take on the benefits of this type of co-location, both for his com-pany and for its partners. Perhaps the most obvious source of synergy: In addition to selling its hardware, Type A also runs a server farm that enables other companies within the hub to manage high-demand periods by doing some or all of their 3D printing liter-ally right down the hall (not only without any shipping
there seems to be violent agreement about the value of collaboration and
communication, and yet we too often give it little more than lip
AnAlytic Solver PlAtformfrom Solver to full-Power Business Analytics in excel
The Excel Solver’s Big Brother Has Everything You Need for Predictive and Prescriptive Analytics
From the developers of the Excel Solver, Analytic Solver Platform makes the world’s best optimization software accessible in Excel. Solve your existing models faster, scale up to large size, and solve new kinds of problems.
From Linear Programming to Stochastic Optimization
Fast linear, quadratic and mixed-integer programming is just the starting point in Analytic Solver Platform. Conic, nonlinear, non-smooth and global optimization are just the next step. Easily incorporate uncertainty and solve with simulation optimization, stochastic programming, and robust optimization – all at your fingertips.
Ultra-Fast Monte Carlo Simulation and Decision Trees
Analytic Solver Platform is also a full-power tool for Monte Carlo simulation and decision analysis, with a Distribution Wizard, 50 distributions, 30 statistics and risk measures, and a wide array of charts and graphs.
Comprehensive Forecasting and Data Mining
Analytic Solver Platform samples data from Excel, PowerPivot, and SQL databases for forecasting and data mining, from time series methods to classification and regression trees, neural networks and association rules. And you can use visual data exploration, cluster analysis and mining on your Monte Carlo simulation results.
Find Out More, Download Your Free Trial Now
Analytic Solver Platform comes with Wizards, Help, User Guides, 90 examples, and unique Active Support that brings live assistance to you right inside Microsoft Excel.
Visit www.solver.com to learn more, register and download a free trial – or email or call us today.
w w w. i n f o r m s . o r g10 | a n a ly t i c s - m aga z i n e . o r g
analyze this!
costs but also without acquiring additional hardware that would likely sit idle during slower times).
In addition, by being physically close to innovators who are using its products, Type A receives constant feedback on where the reliability problems are with their leading-edge machines, what these customers are doing with 3D printing, and what their future needs might be. Similar-ly, by being in close proximity to Type A, the other hub companies have an aware-ness of its product direction, an opportu-nity to get access to early models of new printers, and the chance to collaboratively solve problems as they arise.
For example, Sivertsen shared an in-teresting story of a partner who needed to print something that was officially too big for Type A’s Series 1 printer. Through hallway discussions, the two companies came up with the idea of rotating the de-sign by 45 degrees, after which they man-aged to successfully print a very valuable prototype for the partner right there in Type A’s server farm.
I had originally gone to BAAM looking for an analytics story about capturing and analyzing data to support improved prod-uct reliability, an area where I had done some research [1] in the past. However, for an industry in the very early stages of maturation, more of this type of data is clearly being transmitted in analog rather
than digital format, between people who are deeply knowledgeable and passion-ate about their adjacent/overlapping ar-eas of expertise.
item: After a long career as an opera-tions research leader at P&G [2], Glenn Wegryn is currently a principal at Analyt-ics Impact LLC, as well as president of the Analytics Section of INFORMS. Though I have only met him briefly once, I have been a huge fan of his for a long time. As one of the 1,000-plus members of the Analytics Section, I am thrilled to have someone of his experience and stature leading us. Given his strong track record of successful project outcomes, I definite-ly listen when he speaks [3].
Anyway, I recently came across an email from INFORMS containing We-gryn’s “Top 5 Analytics Predictions for 2015” (see page 14). While the whole list was interesting, No. 4 was the one that caught my eye: “Collaboration and Com-munication (aka the soft skills) will emerge as the difference-maker….”
item: During the last week or so, I have seen a couple of interesting blog posts from Vincent Granville, the creator of Data Science Central (www.datasciencecen-tral.com), billed as “the leading social net-work for big data, business analytics and data science practitioners.” Granville is an
j a n U a r y / f e B r U a r y 2 015 | 11a n a l y t i c s
active and innovative data scientist with a lot of interesting ideas, and I definitely pay attention to what he has to say.
In the first of these blog posts [4], Gran-ville’s suggests that if you were trained as a statistician in the classical sense [as he was], very little of what you have learned is actually all that useful if you are trying to make a living as a data scientist (for the O.R. fans reading this, note that he does state that “data science uses some oper-ations research”). Despite its intentionally
provocative title (“Data science without statistics is possible, even desirable”), this was actually a very thoughtful piece that argued for a less dogmatic and more utilitarian approach to tackling data. The crux of his intellectual objection is most clearly stated near the bottom of the ar-ticle when he writes that, “old statistics use a top-down approach, from model and theory to data, while new statistics or data science use a bottom-up approach, from data to model or algorithm.”
THE COMPETITIVEADVANTAGE DEGREES t eve B .
C lass o f ‘ 15
FOR MORE INFORMATION | analytics.tamu.edu� � �
The Texas A&M program gives me the curriculum I’m looking for, the flexibility to participate remotely and a top-notch staff
of seasoned professors with loads of practical experience beyond the university.
EXCERPT FROM CLASS OF 2015 FIRST SEMESTER EVALUATIONS
MASTERS OF SCIENCE ANALYTICSAT HOUSTON CITYCENTRE OR VIA ONLINE
w w w. i n f o r m s . o r g12 | a n a ly t i c s - m aga z i n e . o r g
analyze this!
The second of these posts [5], fea-tured a list of “10 data science predictions for 2015” originally created by the Institute for Advanced Analytics (http://iianalytics.com). This list also had a heavy empha-sis on collaboration and communication.
So what? There seems to be violent agreement about the value of collabora-tion and communication, and yet we too often give it little more than lip service. Even the term “soft skills” is inherently pe-jorative, the modifier essentially mocking the noun to which it is attached.
I would suggest looking at things a bit differently. Like Sivertsen and his col-leagues at Type A Machines, we as ana-lytics professionals are constantly trying to gain acceptance for new technologies from people with different skills and capa-bilities than the ones who created them. To do this successfully, we must culti-vate our ability to see the world through their eyes and to help them realize how we might be able to help enable them to be successful. These are skills, with no modifier needed.
But if I was required to add an adjec-tive for these skills, the word that comes to mind is “survival.” Because we will always need to be innovating in collaboration with our allies and customers, for in the age of software we can expect solutions to be
commoditized quickly. Lest you imagine otherwise, take note of the last two pre-dictions for 2015 on the IIA list:
• Analytics, machine learning, and cognitive computing will increasingly take over the jobs of knowledge workers.
• Automated decision-making will come of age in 2015.
In other words, innovate or die, my friends. And good luck trying to do it suc-cessfully in isolation. ❙
Vijay Mehrotra ([email protected]) is a professor in the Department of Business Analytics and Information Systems at the University of San Francisco’s School of Management. He is also a longtime member of INFORMS.
RefeRences
1. Mehrotra, v., grossman, t. a., “o.R. process skills transform an out of control call center into a strategic asset,” interfaces, vol. 39, no. 4, July-august 2009, pp. 346-352.
2. see http://www.cbsnews.com/news/how-operations-research-drives-success-at-pg/ for an excellent overview of Wegryn’s work at p&g.
3. for example, check out his excellent presentation from the spring 2012 infoRMs analytics conference entitled “driving competitive advantage with o.R.” at https://www.informs.org/community/analytics/videos/analytics-process-presentations.
j a n U a r y / f e B r U a r y 2 015 | 13a n a l y t i c s
CPLEX Optimization Studio®.Still the best and most complete optimizer and modelerNow 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.
w w w. i n f o r m s . o r g14 | a n a ly t i c s - m aga z i n e . o r g
what’s tRenDing
2014 proved to be a winning year for analytics. Going by the number of conferences devoted to the topic of analytics and attendance, including this year’s record-breaking INFORMS Annual Meeting in San Francisco, all indications are that it will continue to prosper. Here I present five of the most important predictions regarding analytics capabilities for 2015.
1. What’s in a name? There will be continued ef-fort, albeit not fully resolved in 2015, to converge on a common definition of what analytics is. INFORMS’ official definition is “... the scientific process of trans-forming data into insight for making better decisions.” There are numerous other definitions or positions on what analytics is, but a middle ground needs to be found between the data-centric definition of analytics (What can the data tell us?) and the decision-centric version (What is the problem we’re trying to solve?).
Indeed, I view analytics as a bridge to converge the two in peaceful co-existence. Most importantly, it provides an easier point of entry for decision-mak-ers to embrace, organize around, pay for and ulti-mately benefit from all of the tools in the shed. The
Top 5 analytics predictions for 2015
by glenn wegRyn, cap
a middle ground needs to be found between the
data-centric definition of analytics (what can
the data tell us?) and the decision-centric version
(what is the problem we’re trying to solve?).
Glenn Wegryn, president of the Analytics Section of INFORMS and principal at Analytic Impact LLC, offers his top five analytics trends for 2015.
j a n U a r y / f e B r U a r y 2 015 | 15a n a l y t i c s
more we play in the same sandbox, the bigger the castle we can build together.
2. business analytics programs will continue to grow. There are now more than 100 business schools in the United States that have, or have com-mitted to launch, curriculum at the un-dergraduate and graduate levels with degrees or certificates in business ana-lytics [1]. No doubt there will be more in 2015. Clearly the B-Schools have heard the call from McKinsey [2] and others [3] on the significant gap projected be-tween the supply and demand for talent in the analytics space, particularly in the
predict-and-decide advanced analyt-ics skills. In 2015, the more established programs will dig deeper and continue to fine-tune the curriculum while newer programs will close the gap quickly. The more successful programs will lever-age the breadth of academic disciplines (computer science, operations research, engineering, math and statistics, market-ing, finance and others) to strengthen their programs.
3. fraud and security. With the num-ber of security breaches on major corpo-rations being reported almost weekly, such as at Target, The Home Depot and
2015
w w w. i n f o r m s . o r g16 | a n a ly t i c s - m aga z i n e . o r g
what’s tRenDing
more recently at Sony Pictures, there will be a significant increase in investment across the board in safeguarding com-merce and privacy on the Internet. The importance of applying analytics methods – from using decision analysis to guide in-vestment choices, to statistical methods, to detect-and-anticipate breaches and optimization models, to improve infra-structure design for safety, reliability and performance – will accelerate and contin-ue to grow in 2015.
4. collaboration and communica-tion (aka, the soft skills) will emerge as the difference-maker not only in getting the best talent hired into the most coveted roles, but also in enabling more recogni-tion and value to organizations that uti-lize analytics. Soft skills are important to educate, sell the value of, and, ultimately, transform the culture within departments and organizations. But more importantly, these soft skills are essential to clearly convey the context of the problem and to recommend a course of action for a de-cision-maker to take. That can span from intelligent bots interacting effectively with users in an online application, to indepen-dent practitioners seeking to build repeat business with a client, to boardroom ana-lysts able to think on their feet and able to describe in plain language the options available and facilitate a decision.
Communication includes effective use of visual capabilities from Tufte-proud graphics to interactive, data-rich, drill-down tools as a means, and not an end, to a better decision.
5. the infoRMs certified analyt-ics professional (cap®) program will continue to increase in importance as a qualification of skills and expe-rience. Longer term, as the supply of analytic talent catches up to demand, the CAP program will be recognized as an important differentiator for employ-ment. Currently, the numbers who are either certified or are signed up to take the exam exceeds the rate of the Proj-ect Management Professional (PMP) certification, over the same period [4]. To that end, INFORMS should begin a campaign targeted to the “buyer” of analytics to grow awareness of this cer-tification in their hiring decisions. ❙
To hear a podcast of Glenn Wegryn talking about his predictions, visit https://www.informs.org/podcast
RefeRences
1. Research from infoRMs Masters in analytics committee, presented at the 2014 infoRMs annual Meeting.
2. McKinsey institute, “big data: the next frontier for innovation, competition and productivity,” May 2011.
3. informationWeek Reports, “big data Widens analytics talent gap.”
4. presentation, meeting of infoRMs board of directors, nov. 9, 2014.
j a n U a r y / f e B r U a r y 2 015 | 17a n a l y t i c s
Dr. Andrew Jennings, chief analytics officer at FICO, sees a post-big-data world in 2015. Predictive analytics and the big data that fuels it become deeply and broadly em-bedded in business and society – no longer a phenomenon but widely accepted as part of the foundation. Here are a few of his predictions for the coming year:
1. big data and analytics are “busi-ness as usual.” The hype around big data has finally crested. Big data will always be with us, and it will keep getting bigger. But business users have come to the realiza-tion that analytics is the key to unlocking value from big data. In 2015, more compa-nies will cultivate in-house analytic compe-tencies and many will embrace analytics as an important part of their company cultures.
2. predictive security will become an important tool in the effort to stop cyber criminals. With cybercrime and cyber ter-rorism on the rise, the good guys are finally moving beyond their reactive approach to security. In 2015, police, military and in-telligence agencies will employ analytics to predict when, where and how the next attack will occur so such attacks can be thwarted before damage is done.
3. automation of modeling as well as the related reporting will be a priority. The continued expansion of analytics is predicated on making the process as sim-ple as possible and codifying the steps that many analytic experts take for granted. This is not without some danger, but I believe we will see significant investment in this area.
4. the unstructured shall inherit the earth. Analysis of unstructured data is coming of age, particularly in customer ser-vice. Every tweet, chat session, call center conversation, and customer support email is going to be analyzed to accelerate problem resolution, optimize scripts for sales people, enhance shopping experi-ences, make products less confusing, and increase compliance.
5. the causation vs. correlation debate is becoming passé. In most cases whether the underlying relationship is causal or correlative is irrelevant. In business circles, people want to make better decisions. They need to know what business levers to pull to get closer to the desired result; the why is secondary. Overall the important issue is to gather good data and trust it. As famed statistician Bradley Efron said, “Those who ignore statistics are condemned to re-invent it.” ❙
five MoRe pRedictions foR 2015
w w w. i n f o r m s . o r g18 | a n a ly t i c s - m aga z i n e . o r g
healthcaRe analytics
In my last article I reviewed the key achievements, opportunities and challenges facing the healthcare analytics industry in 2014. In this first article of 2015, I will share some of the anticipated developments in the marketplace that will impact the nascent industry.
1. Great year for healthcare analytics professionals
As the demand for healthcare analytics increas-es, the demand for professionals with analytics ex-perience will also grow. Health Catalyst, a healthcare analytics product and consulting company, found in a survey that healthcare analytics is the highest prior-ity IT investment in the healthcare industry followed by population health and ICD-10. A recent report published by Burning Glass Technologies stated the demand for health informatics workers will increase at twice the rate of employment, yet the nation al-ready faces a shortage of qualified candidates. This is good news for health informatics professionals with experience in new business models such as pay-for-performance or value-based purchasing. If you have a blend of those skills and experiences, you will be a hot commodity in 2015.
2. Interoperability will be a key focus area for health IT vendors and the government
Interoperability is still a hot topic within the Of-fice of the National Coordinator for Health Informa-tion Technology (ONC). In an updated release of the
What to expect in 2015
by Rajib ghOsh
the demand for health informatics workers will
increase at twice the rate of employment, yet
the nation already faces a shortage of qualified
candidates.
four predictions for the healthcare analytics industry in the year ahead.
j a n U a r y / f e B r U a r y 2 015 | 19a n a l y t i c s
interoperability roadmap, ONC described one of its key criteria by 2017 is to enable providers to send, receive, find and use a basic set of essential health information. By 2020, the roadmap projects, granular health information will be available to the providers. Interoperability, or lack thereof, is a key obstacle in creating a longitudinal view of a patient and deployment of popu-lation level analytics.
Patients can move from one health system to another during a course of 12 to 18 months. Meanwhile, patient’s medi-cal records get locked in a fragmented
manner inside multiple electronic sys-tems that do not talk to each other. Re-leasing data from such silos is the key to deriving benefits from the use of health-care analytics. Until now, providers, pundits and politicians have blamed elec-tronic health record (EHR) companies for their product strategy of keeping patient data locked into their own proprietary databases. Epic, the leader in market share among EHR companies, recently announced that it has hired a lobbyist firm to fix its “image” on Capitol Hill. Epic argues that its product can connect with
Are You Looking For an Analytics Professional to Make Sense of Your Data?
Find the seasoned professionals you need – over 800 analytics professionals expected
Provide your recruitment materials in a casual setting
Arrange discreet on-site meetings in private booths
Enjoy discounted combination pricing with the fall Annual Meeting Job Fair
Enhance your visibility with an ad in Analytics or OR/MS Today
w w w. i n f o r m s . o r g2 0 | a n a ly t i c s - m aga z i n e . o r g
healthcaRe analytics
many other systems including health information exchanges (HIEs).
Opposition groups argue that Epic can only interoperate with other Epic users; for everyone else there is a steep cost for interoperability. While some argue that fixing their product strategy is a better solution than following a PR strategy, the fact that Epic is doing this suggests that interop-erability will be the topmost agenda for electronic health record companies in 2015.
EHR industry heavyweights including Epic are behind the recent launch of the “Argonaut Project.” It is aimed at delivering some standard proposals around the utility of the new Health Level 7 (HL7) Fast Healthcare Interoperability Resources (FHIR) technology that has the promise of sharing health-care data on the Internet. The real impact of this proj-ect remains to be seen, but it is an encouraging sign.
3. Providers not payers will drive market for analytics
For a long time payers were considered to be the growth driver of healthcare analytics because they have the total cost of care data. Payers can force network providers to share encounter data, which they use to triangulate with cost data to cre-ate meaningful analytics for predictive and pre-emptive population health management. But that perception is changing.
The insurance market in recent times has seen many twists and turns. Opening state- and federal-level insurance exchanges provided opportunities for payers in the form of government business, but it also brought fierce competition. Gaining market share has become the new mantra for the insurance
gaining market share has become the new mantra for
the insurance companies. they are also busy trying
j a n U a r y / f e B r U a r y 2 015 | 21a n a l y t i c s
companies. They are also busy trying to find new business models for the future.
Big providers are under immense pressure to improve their bottom line in the new world of the Affordable Care Act (ACA) and government-driven pay-ment reform. They have to work hard to improve their quality ratings and emer-gency room readmissions to avoid Medi-care penalties charged by the Center for Medicare and Medicaid (CMS). One estimate suggests that readmission pen-alties will cost hospitals $428 million in 2015, up from $227 million in 2014. For hospitals, large and medium, analytics will be their survival strategy. They will drive growth in the healthcare analytics market in 2015, not payers.
4. Care delivery transformation will increase demand for analytics
Healthcare delivery organizations are undergoing rapid vertical and hori-zontal integrations. Health systems are merging, and various care delivery set-tings such as home health and wellness providers are increasingly brought under the umbrella of hospital systems. Care delivery will continue to undergo decen-tralization and transformation. The key to holding such multiple entities together will be the ability to use data and analyt-ics to drive decision process. This trend will continue in 2015 and beyond, creat-ing opportunities for analytics product
companies, along with technology and business consulting service providers.
Many argue that the political landscape change in the November election will halt this transformation in healthcare. I disagree. The train has left the station. Healthcare or-ganizations have already spent hundreds of millions of dollars buying new technol-ogy, investing in their workforce, offering incentives and restructuring their organi-zations in response to the ACA. Millions of people now have obtained insurance and have gotten used to the freedom of buying insurance from a marketplace where insur-ance companies compete. It will be hard, if not impossible, to take all that away.
I’m going to stop worrying about the what-if scenario, which may never ma-terialize, and focus on fixing our broken healthcare system and plugging the loopholes. This is my resolution for 2015. What’s yours? ❙
Rajib Ghosh ([email protected]) 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. Follow Ghosh on twitter @ghosh_r.
w w w. i n f o r m s . o r g22 | a n a ly t i c s - m aga z i n e . o r g
When you have an ad hoc analytical project, internal IT dependency usually comes your way. Until a few years back, we had to accommodate and account for it in our delivery timelines. In today’s world, we can avoid this altogether by opting for public cloud analytics.
thE PRoBlEm
Let’s assume that you are asked to analyze a sig-nificantly large data set and provide insights. Based on the data size, you determine that you’d require a clustered and distributed environment to be able to run the analysis within the pre-decided timeline. You ask your manager for help and he says to raise a re-quest with IT. You raise the request in the ticketing system and check with IT. They answer that they don’t have it ready and will need to procure the system first. You comply and ask how long will the process take? The IT personnel responds, “Four weeks provided you get the budget approval from your department.” Your timelines have just run amok!
If the above scenario is true in your case, then the situation is similar to what I have heard from several
j a n U a r y / f e B r U a r y 2 015 | 23a n a l y t i c s
other analysts in different organizations. While the IT department centrally man-ages all infrastructure, they mostly focus on normal operational efficiency goals. Project specific ad hoc requirements usually do not fall under the preview of IT supply and demand hence is looked upon as ad-hoc. This is the norm in every service industry today. The 30,000-foot question is: What do we do?
ClouD analytiCs to thE REsCuE
The obvious answer comes in the form of using public cloud infrastructure
for your analytical needs. Using public cloud practically eliminates the time lapse on infrastructure procurement. In addition, analytics on public cloud comes in three different flavors:
1. The black-box approach allows you to consume already trained models as services;
2. The white box approach allows you to use the public cloud infrastructure for data storage, deploy analytical tools for performing descriptive or predictive needs and a visualization platform for re-porting insights; and
BE A PART OF THE ANALYTICS REVOLUTION!
Join the World of Certified Analytics Professionals (CAP®)
®
Set Yourself Apart. Join the Analytics Revolution. Become a CAP® Now. Visit http://www.informs.org/certification or email us at [email protected] with any questions.
w w w. i n f o r m s . o r g24 | a n a ly t i c s - m aga z i n e . o r g
viewpOint
3. The grey box approach allows you to create an ensemble of models using sta-tistical techniques such as decision trees and infer underlying rules for relationships.
In this case, while the actual imple-mentation of statistical technique is black boxed, you still have control in deter-mining the right path and relationships by varying the factors required for rules determination. All you need is to make a choice of determining which approach to take based on the problem definition.
Custom analytiCs on
PuBliC ClouD
As part of an R&D engagement for one of our clients, we needed to perform sentiment analysis on tweets in order to determine the problem context, key-words, sentiments and topics. The main objective of this exercise was to deter-mine service-related issues and result-ing sentiment of customers. This being a social media analytics exercise, the solu-tion was to incorporate aspects of real-time sentiment complex event processing along with data aggregation and historical descriptive analytics.
We determined that using a Twitter streaming service would be the best op-tion. This would require constantly listen-ing to the streaming service and would result in collecting and analyzing up to 2,000 tweets a minute.
Based on the infrastructure require-ment, we looked into public cloud pro-viders options and decided to go with Amazon Web Services. It took us just a couple of days (includes internal approval process) to get three large capacity nodes with Linux pre-installed. We wanted a clean system as we were building a cus-tom natural language processing system to process the tweets.
It took us one week to deploy Apache Hadoop, Druid, R statistical tool, Zero MQ, Python and Java and two weeks to run through performance benchmark test-ing and an optimization exercise. In par-allel, we also developed and deployed a custom HTML5 based visualization tool for insights.
All in all, the entire exercise took a month and a half from start to finish for the initial proof of concept. The NLP algo-rithm development journey is still under-way and being constantly optimized for accuracy and performance.
What could have easily been a three months exercise if we had waited for internal infrastructure procurement, took only half the time to deliver. We also have both data ingestion and insights visualization on the public cloud along with backend processing components – not to mention, we did not make any depreciating investments, hence keeping the bottom line intact.
j a n U a r y / f e B r U a r y 2 015 | 25a n a l y t i c s
ConClusion
Cloud-based analytics or cloud analytics is an evolution in the making. We are seeing traditional BI vendors pro-viding cloud-based SAAS-BI services on one hand, while the others are provid-ing specific niche services in social me-dia analytics using third-party sourced data. In between, we have big cloud in-frastructure providers, providing both in-frastructure and analytical platform. On the most crucial aspects of data security, most of them have ironclad service-level
agreements and adhere to them. Based on our own experience and listening from other engagements, I can confidently say, “Cloud analytics works for me,” provided you know how to make it work for you. ❙
Ganesh Moorthy is an associate director at Mu Sigma, where he serves as program manager/senior solution architect for R&D engagements. Moorthy has more than 16 years of experience in leading enterprise solution development for Fortune 500 clients. He is currently involved in building industrial Internet, augment reality and analytics and visualization platforms for both descriptive and predictive analytics.
INFORMS Continuing Education program offers intensive, two-day in-person courses providing analyticsprofessionals with key skills, tools, and methods that can be implemented immediately in their work environment.
These courses will give participants hands-on practice in handling real data types, real business problems and practical methods for delivering business-useful results.
INTRODUCTION TO MONTE CARLO AND DISCRETE-EVENT SIMULATIONLearn the basics of Monte Carlo and discrete-event simulation, how to identify real-world problem types appropriate for simulation, and develop skills and intuition for applying Monte Carlo and discrete-event simulation techniques.May 28-29, 2015 - Washington D.C.
COURSES FOR ANALYTICS PROFESSIONALSducation
continuing Visit the Website for the latest course schedule: informs.org/continuinged
FOUNDATIONS OF MODERN PREDICTIVE ANALYTICSLearn data mining techniques and tools that will allow you to make the link between business needs and your technical skills.April 15-16, 2015 - Huntington Beach, CA
ESSENTIAL PRACTICE SKILLS FOR ANALYTICS PROFESSIONALSLearn practical frameworks and systematic processes for addressing complex, real-world problems and how to facilitate effective action.February 26-27, 2015 - San Jose, CA
DATA EXPLORATIONAND VISUALIZATIONLearn methodologies, processes and essential tools for exploring and visualizing data in order to derive insights and knowledge.March 19-20, 2015 - Dallas, TX
w w w. i n f o r m s . o r g26 | a n a ly t i c s - m aga z i n e . o r g
Certification has roots in medieval craft guilds, which established standards, set ethical guidelines for practice and functioned as professional associa-tions. The Institute for Operations Research and the Management Sciences (INFORMS) is the largest pro-fessional society in the world for analytics, operations research and management science. INFORMS’ Certi-fied Analytics Professional (CAP®) Program is an effort to establish standards and set guidelines. Certainly the practice of analytics is a craft, requiring special skills in its execution. Let me explain, and maybe by the end you’ll be interested in getting certified yourself.
PRaCtiCing youR CRaFt in thE miDDlE agEs
In the high Middle Ages, learning a craft began with apprenticeship, whereby teenage boys worked for a master craftsman in exchange for room, board and training. Apprentices could only practice their craft in their master’s workshop until they became journeymen, which required several years of expe-rience and the production of a high-quality piece of work in that craft. However, once reaching journey-man status, skills became portable, an journeymen were granted letters or certificates that allowed them
CAP: How to become a modern-day analytics journeyman
j a n U a r y / f e B r U a r y 2 015 | 27a n a l y t i c s
to travel to other towns to practice their craft, earn wages and learn from other masters.
After more years of practice and pro-duction of a masterpiece the journey-man becomes a master, if other masters agreed. The new master was then free to set up his own workshop. Guilds promot-ed their craft, set its norms or standards, and maintained a code of ethics related to its practice. They offered practitioners a portable credential that could signal their skills to those who wanted to do more
than just take their word. The travels of the journeymen had the added effect of disseminating new ways of practice, since they brought their accumulated knowl-edge to each new location and workshop.
inFoRms, a moDERn-Day guilD?
INFORMS is certainly not a medieval guild, but it is a global organization of 11,000 members, of whom 50 percent are academics, 30 percent practitioners and 20 percent students. Members are highly educated – 96 percent have or
Winner: Identifying Risks and Mitigating Disruptions in the Automotive Supply Chain Project Portfolio Planning at Intel Corporation Airline Crew Augmentation, Decades of Improvements from Sabre Vaccine Prioritization for Effective Pandemic Response
Statistical and Optimization Techniques for Laundry Portfolio Optimization at P&G Gerrymandering for Justice: Redistricting U.S. Liver Allocation
Videos of 2014 Wagner Prize presentations are now available at:http://livewebcast.net/INFORMS_AM_Wagner_Prize_2014/pres1/
2014 Wagner Prize winners Oleg Gusikhin (L) and David Simchi-Levi (R) with Wagner Prize Chair Allen Butler (C).
w w w. i n f o r m s . o r g28 | a n a ly t i c s - m aga z i n e . o r g
fORum
are earning an advanced degree, and 51 percent have Ph.D.s. INFORMS has taken on the championing of analytics in many ways, including establishing con-tinuing education programs; an analyt-ics maturity model; engagement with industry, government and academia; and certification.
The Institute’s decision to establish a certification program shares goals with the guilds: to advance the craft by intro-ducing standards of quality, identify indi-viduals with breadth of knowledge and encourage continued competency. While guilds were quasi-governmental in re-quiring membership to practice that craft (akin to governments requiring licenses for certain professions today), certifica-tion is a voluntary credential granted by a non-governmental body.
INFORMS conducted a job task anal-ysis study by a panel of subject matter ex-perts to define the knowledge, skills and abilities necessary to effectively practice analytics. They identified seven domains: business problem framing, analytics problem framing, data, methodology se-lection, model building, deployment, and model lifecycle management.
What is REquiRED FoR
CERtiFiCation?
There are five “Es” of the Certified Analytics Professional (CAP®) program:
education, experience, exam, effective-ness and ethics. The education and analytics-related work experience com-bination required is three years for those with a master’s degree (any guess where that term originated?) in a related disci-pline (e.g., statistics, mathematics, com-puter science, etc.), five years with a bachelor’s degree in a related discipline and seven years with a degree in an unre-lated discipline (waivers of the educational eligibility requirements will be considered on a case by case basis). The exam tests skills and knowledge in the seven JTA do-mains with 100 multiple-choice questions that must be completed in three hours or less. Computer-based testing is available at 700+ locations worldwide.
Analytics requires more than technical skills and knowledge, so a confirmation of effectiveness in the area of soft skills by an employer or client is required. And, CAP requires agreement with the code of ethics. For more details on applying see the CAP Candidate Handbook and the Complete CAP Study Guide.
Why gEt CERtiFiED?
The target group is early to mid-career professionals, or “apprentices” seeking to become “journeymen,” although thank-fully today apprentices are paid! Those advanced in their careers may have a re-sume whose mastery stands on its own,
j a n U a r y / f e B r U a r y 2 015 | 29a n a l y t i c s
but early career professionals may find the differentiation offered by a portable, independent, third-party validation of knowledge of certain standards helpful. Not unlike the journeyman certificate or letter, in this era of increasing demand for analytics professionals certification attests to knowledge and skills beyond a person’s own word. It demonstrates commitment to career and to craft and adherence to accepted standards. It can increase earnings potential and provide the personal satisfaction of achieving a milestone. INFORMS and CAP benefits
will endure for years to come.For more on the CAP program,
click here. ❙
Polly Mitchell-Guthrie ([email protected]) is the senior manager of the Advanced Analytics Customer Liaison Group in SAS’ Research and Development Division, where her team serves as a bridge between R&D and external customers and internal SAS divisions. She serves as vice chair of the INFORMS Analytics Certification Board and has an MBA from the Kenan-Flagler Business School at the University of North Carolina at Chapel Hill. She is a member of INFORMS.A version of this article appeared in the December 2014 issue of Swiss Analytics Magazine. Reprinted with permission.
https://analyticsmaturity.informs.org
· Assess the way your organization uses analytics
· Create an improvement plan
· Track your progress
· Tap the resources of the leading professional association for advanced analytics professionals
THE INFORMS DIFFERENCE
· Reputable
· International Leader
· Free Access to INFORMS Analytics Maturity Model
· Resilient Model That Grows to Meet Your Needs Begin using the beta version of the INFORMS Analytics Maturity Model. Just go to https://analyticsmaturity.informs.org. Start today! For additional information, email [email protected] or call 433-757-3500, ext. 560
w w w. i n f o r m s . o r g3 0 | a n a ly t i c s - m aga z i n e . o r g
INFORMS recently announced six finalists repre-senting a diverse collection of applied analytics and advanced operations research projects that will com-pete for the 2015 Franz Edelman Award in a series of judged presentations at the INFORMS Conference on Business Analytics & Operations Research in Hun-tington Beach, Calif., in April.
Widely considered the “Super Bowl” of opera-tions research, as well as INFORMS’ highest organi-zational honor, the Edelman acknowledges the best application of high-end analytics. The winner will be announced at an evening Edelman Gala held in conjunction with the conference.
The finalists and their presentations include:• IBM for “Predictive Cloud Computing with Big
Data: Professional Golf and Tennis Forecasting” • Ingram Micro for “End-to-End Business Analytics
and Optimization in Ingram Micro’s Two-Tier Distribu-tion Business”
widely considered the “super bowl” of
operations research, as well as infORms’
highest organizational honor, the edelman award
acknowledges the best application of high-end
analytics.
infORms init iatives
Edelman Award, continuing education, IAA finalistsinfORms announces six finalists for the 2015 edelman award
j a n U a r y / f e B r U a r y 2 015 | 31a n a l y t i c s
• LMI/Defense Logistics Agency for “Peak and Next Gen: Effective Inven-tory Control for Items with Infrequent or Frequent, Highly Variable Demand”
• Saudi Arabia Ministry of Municipal and Rural Affairs for “A Pilgrim Schedul-ing Approach to Increase Public Safety during Hajj”
• Syngenta for “Good Growth through Advanced Analytics”
• U.S. Army/Sandia National Labora-tories for “The Capability Portfolio Analy-sis Tool (CPAT): A Fleet Management and Decision Analysis Tool”
Now in its 44th year, the Franz Edel-man Prize competition honors outstand-ing examples of analytics and operations research projects that transform compa-nies, entire industries and people’s lives. Using innovative advanced analytics methods, the nominated teams were in-strumental in helping their respective in-stitutes make better decisions, providing a disciplined way to which management can improve organizational performance in a wide variety of situations and across both public and private organizations.
For more information, click here.
2015 sChEDulE oF Continuing ED
analytiCs CouRsEs
INFORMS’ continuing education pro-gram is designed specifically for those in the field of analytics and operations
research. INFORMS, the world’s largest organization for advanced analytics pro-fessionals, produces and disseminates data analytics and operations research information, training courses, certifica-tion, and best-in-class resources for our members and the industry at large.
Four upcoming intensive hands-on courses provide real take-away value. You will leave these courses with the skills, tools and methods that you can implement immediately in your work. Included are two of the newest courses: Introduction to Monte Carlo and Discrete- Event Simulation and Foundations of Modern Predictive Analytics. Following are the dates and locations for the 2015 February through May courses:
Essential Practice Skills for Ana-lytics Professionals, Feb. 26-27, San Jose, Calif.
Learn how to integrate and apply your analytical skills to real-world problem-solving for businesses and other organi-zations. This course provides approaches that you can apply immediately to a wide variety of settings, whether within your own organization or for an external client.
Data Exploration & Visualization, March 19-20, Dallas
Hands-on training that focuses on the critical steps in the process of analyz-ing data: accessing and extracting data, cleaning and preparing data, exploring
w w w. i n f o r m s . o r g32 | a n a ly t i c s - m aga z i n e . o r g
infORms init iatives
and visualizing data. This course uses several of the most popular software tools intensively, and provides an overview of the range of available software options.
Foundations of Modern Predictive Analytics, April 15-16, Huntington Beach, Calif.
Learn data mining techniques and tools that will allow you to make the link between business needs and your techni-cal skills. This course gives you hands-on practice in handling real data types, real business problems and practical methods for delivering business-useful results.
Introduction to Monte Carlo and Discrete-Event Simulation, May 28-29, Washington, D.C.
Identify real-world problem types ap-propriate for simulation, and develop skills and intuition for applying Monte Carlo and discrete-event simulation techniques. Via hands-on interactive sessions, you will investigate the use of Monte Carlo simu-lation in decision-making and the use of discrete-event simulation to solve mathe-matically intractable problems in stochas-tic modeling.
Don’t miss out on these courses. For detailed course information on all INFORMS courses and to register, visit www.informs.org/continuinged or contact Thedra White at thedra.white@ informs.org or 443-757-3570.
innoVatiVE aPPliCations in
analytiCs aWaRD Finalists
The Analytics Section of INFORMS has named three finalists for its Innovative Applications in Analytics (IAA) Award. The Section’s flagship prize, the IAA recog-nizes creative and unique developments, applications or combinations of analytical techniques used in practice.
The three finalists, chosen by a panel of judges following a two-month verifica-tion process, will present their work at the INFORMS Conference on Business Analytics and Operations Research in Huntington Beach, Calif., in April 2015, where the winner will be announced.
The three finalists include:• “Simulation Approach for Aircraft
Spare Engines & Engine Part Planning,” Jose A. Ramirez-Hernandez, Steven Oakley, Mei Zhang, Alejandro Scalise, American Airlines
• “Intelligent Surgical Scheduling System,” Narges Hosseini, Kalyan Pasupathy, Yariv Marmor, Thomas Rohleder, Jeanne Huddleston, Paul Huddleston, Mayo Clinic Rochester
• “Fusion Analytics for Public Trans-port Event Response,” IBM Research and the Land Transport Authority of Singapore
Solve key business problems utilizing big data. Earn an AACSB-International accredited Master of Business Administration with a specialization in Business Analytics from the University of South Dakota.
Learn more: www.usd.edu/cde
The University of South Dakota’s Beacom School of Business has been continuously accredited by AACSB-International since 1949.
Advance your career with an online Master of Business Administration with a specialization in Business Analytics.
DIVISION OF CONTINUING & DISTANCE EDUCATION414 East Clark Street | Vermillion, SD 57069605-677-6240 | 800-233-7937www.usd.edu/cde | [email protected]
C
M
Y
CM
MY
CY
CMY
K
USD_Online MBA BA Analytics Magazine Ad.pdf 1 6/9/14 9:15 AM
w w w. i n f o r m s . o r g34 | a n a ly t i c s - m aga z i n e . o r g
newsmakeRs
inFoRms soCiEty namEs ‘Rigs to REEFs’
toP solution
Lumina Decision Systems, led by CEO Max Henrion, won the 2014 Decision Analysis Prac-tice Award co-sponsored by the Decision Analysis Society of INFORMS and the Society of Decision Professionals for the project titled “A Multi-attribute Decision Analysis for Decommissioning California’s Offshore Oil Platforms.” Co-authored by Brock Bern-stein and Syya Swamy, Henrion delivered the award-winning presentation at the 2014 INFORMS Annual Meeting in San Francisco in November.
The award recognizes outstanding applications of decision analysis to significant decisions. It considers the quality of the analysis, its impact on the decision, and the importance and benefit of that decision.
“This was a great demonstration of how decision analysis can bring together parties with significantly dif-ferent points of view and objectives to find a ‘win-win’ so-lution,” says Frank Koch, chair of the award committee.
Like many environmental decisions, the ques-tion of how to decommission California’s offshore oil platforms started out as highly controversial. But, re-markably, almost all stakeholders, including oil com-panies and environmentalists, ended up supporting the same solution. The original leases required the oil companies to remove the entire rigs at the end
Lumina, Chevron earn DA honorslike many environmental
decisions, the question of how to decommission california’s offshore oil
platforms started out as highly controversial.
but, remarkably, almost all stakeholders, including
oil companies and environmentalists, ended up supporting the same
j a n U a r y / f e B r U a r y 2 015 | 35a n a l y t i c s
of their productive life. Some are up to 1,200 feet deep, making removal very costly with substantial environmental impacts. These rigs have developed their own rich ecosystems, with commer-cially and ecologically important rockfish species, making them popular with sea lions, recreational divers and other ma-rine life forms.
An interdisciplinary team created a decision tool in Lumina’s Analytica software to explore and evaluate de-commissioning options to assist the California Ocean Science Trust. The two
main options examined were complete removal, as required by the original leas-es, and conversion to an artificial reef, by cutting structures at 85 feet below the water line to prevent interference with shipping. Partial removal reduces costs by up to $500 million and helps to pre-serve marine ecosystems. Eventually the stakeholders reached near consensus on the “rigs to reefs” policy option. Cali-fornia lawmakers almost unanimously passed enabling legislation, signed into law by then-Gov. Arnold Schwarzeneg-ger in 2010.
“By clearly identifying the issues, synthesizing the best multi-disciplinary science, daylighting the uncertainty and providing unbiased re-view, the tool was success-ful in distilling the rhetoric to meaningful discussion of trade offs and values,” says Skyli McAfee, director of Cali-fornia Ocean Science Trust. “Further, the tool was made available to the public; its assumptions and approach were transparent. Constitu-ents had the opportunity to import various scenarios and learn the best approach. With this tool, sound legislation was passed that will serve California and our marine resources well.”
“This decision analysis was one of the most challenging and, in the end, satisfying projects I have worked on,” adds Henrion. “It is truly an honor to have our work recognized in this way.”
ChEVRon RECEiVEs inauguRal
RaiFFa-hoWaRD aWaRD
The Society of Decision Profession-als (SDP) presented the inaugural Raiffa-Howard Award for Organizational Decision Quality to the Chevron Corporation during
a gala celebration mark-ing the 50th anniversary of decision analysis. The cel-ebration, which included an afternoon program on the history and future of decision analysis followed by an evening dinner gala, was held Nov. 8 in San Francisco in conjunction with the 2014 INFORMS Annual Meeting.
The award recognizes “consistent and sustained excellence in decision-making throughout an organization.”
Chevron was cited for its “unique and long-term commitment to making de-cision quality an organi-
zational competency contributing to the company’s outperforming its peer group in total shareholder return for both the last five- and 10-year periods. Chevron began its decision analysis initiative more than 15 years ago. Today, more than 25,000 employees have been trained in the dis-cipline. Decision analysis is the discipline incorporating decision theory, method-ology and professional practice guiding decision-making under uncertainty.”
The awards’ namesakes – Profes-sor Howard Raiffa of Harvard University
j a n U a r y / f e B r U a r y 2 015 | 37a n a l y t i c s
and Professor Ronald Howard of Stan-ford University – are widely considered co-founders of the discipline of decision analysis, which crystallized with key contributions from Raiffa and Howard in 1964. Both were on hand for the cel-ebration, along with more than 300 other decision analysis luminaries and guests.
“While elements of decision analysis have existed for hundreds of years, it was the work of [Raiffa and Howard] beginning in 1964 that brought it into mainstream use by giving us a logical and systematic set of principles to enhance our clarity of thought and enable us to act with confi-dence in the face of uncertainty,” says Bill Klimack, president of the SDP, an associ-ation of professional practitioners. “Corpo-rations and government entities now have the tools to make complex decisions.”
“Decision analysis has by now helped many thousands of decision-makers, and has created many billions of dollars of value for the economy through better business decisions and better govern-ment policies,” adds Jeff Keisler, presi-dent of the Decision Analysis Society (DAS), a subdivision of INFORMS.
“The Society of Decision Profes-sionals is delighted to honor professors Howard and Raiffa for their work that has made an indelible impression on our field and our lives,” adds Jack Kloeber, vice president of SDP.
Klimack, Keisler and Kloeber are ac-tive members of INFORMS as well as SDP. INFORMS was a sponsor of the gala celebration along with DAS.
Decision analysis emerged as a dis-tinct academic field in 1964, building on developments in statistical decision the-ory and game theory by Raiffa and dy-namic probabilistic systems by Howard. In that year, Raiffa began teaching the first university course in decision analy-sis within the Economics Department at Harvard and began preparing material for his 1968 book, “Decision Analysis.”
Also in 1964, but independently, Howard conducted the first professional application of the field he called “deci-sion analysis,” which he described in his 1966 paper, “Decision Analysis: Applied Decision Theory.” Since then, major cor-porations have adopted decision analy-sis to improve multibillion-dollar capital decisions. ❙
Source: Raiffa-Howard Award Program
help promote Analytics MagazineIt’s fast and it’s easy! Visit: http://analytics.informs.org/button.html
w w w. i n f o r m s . o r g38 | a n a ly t i c s - m aga z i n e . o r g
eXecutive eDge
mal l - and medium-s ize businesses (SMBs) are no different than their larger counterparts in recognizing
the growing importance of data analyt-ics. Some view the cost as prohibitive because data analytics, and specifically predictive analytics, is rooted in statistics and sophisticated mathematics that re-quire special skills or expensive outsourc-ing. Others are reluctant to embrace data analytics because they question whether
data can be strategically analyzed to create actionable responses rather than fishing expeditions that may or may not yield anything of value.
The business reality is that these posi-tions have very little to do with data, data science or technology. It is a lack of ex-ecutive guidance and direction – in other words, a people problem. With strong corporate leadership and managerial judgment, an effective analytics strategy is likely to emerge.
Leadership: The real game changer for using data analytics
j a n U a r y / f e B r U a r y 2 015 | 39a n a l y t i c s
The best-known example of this type of leadership is graphically displayed in the film “Moneyball.” Analyzing data about ball-player skills that other general managers overlooked enabled Oakland’s Billy Beane to become one of the first baseball execu-tives to leverage its value in a much more stringent, scientific way – a classic exam-ple of motivational leadership in analytics.
lEaDERshiP’s Data usaBility
Cost has and always will be a factor in analytics decision-making, but thanks to decades of technological advancements, data analytics are now available at prices that small and medium-sized businesses should be willing to consider for the most obvious and important reason: posi-tive impact on the bottom line. A report published by The Economist describes a “strong link” between a company’s fi-nancial success and its use of analytics. A survey released by the publication not-ed that 53 percent of respondents in its strategic group reported a higher finan-cial success rate than their peers who did not take advantage of strategic data management (Briody, 2011). The report indicated that what it called “data collec-tors” (those who do not fully leverage the data at their disposal) and “data wasters” (under-users of collected data) are at risk of coming out on the short end financially.
There is, however, one other obstacle
thE analytiCs quanDaRy FoR
lEaDERs
An analytical focus can be an ultimate game changer for companies, which is why the concentration should be on es-tablished objectives rather than amass-ing data that may or may not be relevant. It is at this point that leadership finds itself at a crossroads of two conflicting scenar-ios, both of which have drawbacks.
One scenario takes the form of a “wait and see” approach before deciding if in-vestment in analytics offers an acceptable return on investment – a risky approach. Delays can impact marketplace position-ing and create an edge for competitors. The second scenario focuses on the here and now of tactical day-to-day operations in lieu of a broader business strategy that leverages data.
An underlying assumption on the part of upper management that applies to both scenarios is erroneous. The assumption: Analytics is only possible with big and costly infrastructures that require signifi-cant management and resource commit-ments. The truth is significantly different. While data is certainly a huge entity, size in this case is not as important as determin-ing relevance. The idea is to keep matters manageable. The best way to do that is to focus on the low-hanging fruit of analytics – the data that is most easily accessible and valuable for establishing achievable goals.
w w w. i n f o r m s . o r g4 0 | a n a ly t i c s - m aga z i n e . o r g
eXecutive eDge
that is not so readily apparent. “The biggest bar-rier is the assumption that analytics is a data sci-ence and technology problem when it’s really a leadership and managerial problem,” says Florian Zettelmeyer, director of the data analytics program at the Kellogg School of Management, Northwest-ern University. “The managerial class has to un-derstand the problems that need to be solved and how analytics can help solve them.”
When leadership fails to implement a decision management strategy and convey its importance throughout the company, it unwittingly bypasses a major growth opportunity. Data constantly flows into and around an organization. Leveraging its specific components requires analytical execution and deployment in areas of maximum benefit by a committed staff. This is the essence of oppor-tunity management. Effective decision manage-ment uses analytics to achieve better outcomes from customer interactions. The value, of course, depends on how well all these factors are applied.
RECognizE thE unRECognizaBlE
Even limited analysis of minimal levels of data can make a significant difference to the bottom line, especially when leadership clearly establishes goals beforehand. The objective for data analyt-ics is not quantity of data but quality of outcomes: e.g., solving business problems, generating cus-tomer information or analyzing market conditions. Results should be sustainable, which means plac-ing the emphasis on strategic instead of tactical. Analytical approaches tend to vary depending upon
even limited analysis of minimal levels of data can make a significant
j a n U a r y / f e B r U a r y 2 015 | 41a n a l y t i c s
directly relates to company goals. There is no reason to go data fishing in the hopes of snagging information that may be vital someday. This isn’t to suggest that such effort has no value. Rather, it’s more about priority, available time and resources.
ClaRiFy Roi anD analytiCal
oPtions
Sometimes return on investment from automated analytical software is difficult to establish, which is why it can be the biggest barrier to a go/no-go decision. Nonetheless, analytics success stories are everywhere. Leslie Deich, a profes-sional financial services specialist, says she found the technology to be extreme-ly effective in detecting consumer fraud, which helped protect GE Consumer Capital, Genworth Financial and Fannie Mae from what could have been signifi-cant losses. “We needed to find fraud patterns in the data and this made every-thing easier because it was able to locate the patterns in very quick fashion,” Deich says. “If you don’t know the data or are unfamiliar with it, this will help you see the relationships.”
CFCU’s David Hufnagel relies on op-portunity management “to get a 360-degree view” of its membership. “It’s meant better targeting (and) now we’re seeing greater acceptance of offers,” he says. “It allows for greater efficiency with each campaign.”
the size of the company. Larger corpora-tions generally engage statisticians in ad-dition to a business intelligence team as part of their overall strategy. Most SMBs cannot afford to contract with statisticians, but have other avenues to maximize busi-ness intelligence, most notably automated programs. Some add BI expertise to their staffs, but are uncertain what their next big step should be. At this stage, leadership has to be prepared to resolve inevitable internal conflicts about analytical direction.
What is needed is a strategy to inte-grate analytics into decision manage-ment beginning with two very simple and achievable practices: focus and common sense. “You must have a long-term plan and a strategy that works with it,” says David Hufnagel, chief operations officer and vice president of Congressional Fed-eral Credit Union (CFCU). “If you don’t work in concert with the plan, it becomes inoperable with no results.” The credit union has relied on automated analytics for the last three years to drive a proac-tive outreach to its 45,000 members. “It’s not about being unique with each individ-ual,” Hufnagel says. “It’s about identifying what is relevant to our members.”
The CFCU experience is an example of focus and common sense for maximiz-ing the value of data analytics through technological innovation. The idea is to have the technology focus on data that
w w w. i n f o r m s . o r g42 | a n a ly t i c s - m aga z i n e . o r g
These experiences demonstrate the ability of companies to achieve their desired outcomes through programs that maximize qualitative and quantitative analytics without having to worry about programming and algorithms. “Analytics are not dis-embodied truth,” says Northwestern’s Zettelmeyer. “They answer those questions managers are most concerned about.”
PathWay to oPPoRtunity
Leaders who take a strategic and proactive ini-tiative when it comes to data analytics set the stage for the successful and actionable use of informa-tion. The starting point for this type of initiative is a focus on management and oversight.
Analytics may be among the most powerful tools for executive decision-making and business manage-ment. They are at their best when flexible, focused and applied in tandem with decision management best practices that can be augmented and enhanced as needed. Executives and managers have a path-way to opportunity management regardless of corpo-rate size and the ability to streamline business flow to improve growth opportunities and increase profitabil-ity if they are willing to take advantage of it. ❙
Bill Gossman ([email protected]) is president of Advanced Software Applications (ASA Corp.) in Pittsburgh, Pa. ASA provides analytical and decision solutions that help businesses grow revenue, improve efficiencies and mitigate risks.
RefeRences
briody, dan, 2011, “big data: Harnessing a game-changing asset,” a report by the economist intelligence unit, sponsored by sas.
eXecutive eDge
leaders who take a strategic and proactive initiative when it comes
to data analytics set the stage for the successful
PRESENT A POSTER AND SAVEPresent your work at the conference known for real-world analytics and save on registration. Case studies, best practice examples, and academic research with a practitioner orientation are all welcome.
LEARN HOW ANALYTICS AND O.R. CAN MAXIMIZE THE VALUE OF YOUR DATA TO DRIVE BETTER BUSINESS DECISIONS. √ Real-world use of descriptive, predictive, and prescriptive analytics
√ Focus on big data, marketing analytics, healthcare applications
√ Most rigorous and real world analytics conference offered
√ Administration of Certified Analytics Professional (CAP®) exam
√ Not just what to do but HOW TO DO IT Only INFORMS can provide an analytics and O.R. conference backed up by the best minds in industry and academia. Hand-picked speakers take you through case studies on how analytics can maximize the value of your data, driving better business decisions, and impacting the bottom line.
http://meetings.informs.org/analytics2015
KEYNOTE: THE INDUSTRIAL INTERNETMr. William Ruh is vice president of GE Software, an advanced software technology and shared services organization that serves more than 30,000 engineers across GE business divisions, giving customers the capabilities required to connect intelligent machines, data and people, and to run industrial-scale analytics.
w w w. i n f o r m s . o r g44 | a n a ly t i c s - m aga z i n e . o r g
ata visualization is enter-ing a new era. Emerging sources of intelligence, theoretical developments
and advances in multidimensional imag-ing are reshaping the potential value that analytics and insights can provide, with visualization playing a key role. The prin-ciples of effective data visualization won’t change. However, nextgen technolo-gies and evolving cognitive frameworks are opening new horizons, moving data visualization from art to science.
Looking back, much attention has been given to the principles of effective data visualization, such as substance, context and actionability. As timeless te-nets that will continue to be important, regardless of medium or format, a brief review seems in order:
• As with any form of communication, effectively conveying a message with data requires that it be substantive. And while creative visuals can enhance interest and memory, embellishment can’t make up for lack of substance. According to purist
j a n U a r y / f e B r U a r y 2 015 | 45a n a l y t i c s
increasingly important to organizations, generating actionable insights remains a top challenge. Learn more.
Looking forward, nextgen technolo-gies and evolving cognitive frameworks will boost the role that data visualization can play in organizations and society. Consider the Internet of Things, Net-work and Complexity Theories, and re-cent developments in multidimensional visualization:
The Internet has transformed the way we visualize information through a better understanding of networks and an explo-sion in profile, behavioral and attitudinal data. Sociograms, for example, have gone from relatively simple graphs to mul-tifaceted relational maps, as illustrated
Edward Tufte, “Every single pixel should testify directly to content.” Learn more.
• Visualization should be accurate and contextual. David McCandless’s Bil-lion Dollar O’Gram provides an example of how greater meaning can be added by incorporating the bigger picture. Accord-ing to McCandless, “Absolute figures in a connected world don’t give you the whole picture. They’re not as true as they could be. We need relative figures that are con-nected to other data so that we can see a fuller picture.” Learn more.
• More than anything else, data visu-alization should facilitate decision-mak-ing, a goal that is difficult to achieve for many. According to a recent KPMG study, while data and analytics are deemed
Source: David McCandless http://informationisbeautiful.net
w w w. i n f o r m s . o r g4 6 | a n a ly t i c s - m aga z i n e . o r g
Data visualizatiOn
figure 2: Pre-Internet sociogram.
figure 3: Internet-age sociogram.Source: Leadership Learning Community
Source: Journal of Social Structure
in Figure 2 and Figure 3, courtesy of the Journal of Social Structure and the Leadership Learning Community.
The Internet of Things is expected to have a similar impact, with billions of connected devices capturing human and machine activity. Fully capitalizing on the data generated will require further ad-vances in our ability to synthesize and display spatiotemporal activities.
Network Theory has been in use for de-cades, with its earliest applications largely in social structure analysis. More recently,
j a n U a r y / f e B r U a r y 2 015 | 4 7a n a l y t i c s
intuitive understanding of these process-es, to distinguish relevant from peripheral factors, to predict their time course, and to locate their origin. However, we show that complex spatiotemporal patterns can be reduced to surprisingly simple, homo-geneous wave propagation patterns.”
By providing insight into the workings of dynamic systems with interdependent elements, Complexity Theory can help us identify trends that might lead to unex-pected tipping points such as environmen-tal disasters. Complexity Theory can also support the decoding of intricate structur-al dependencies such as economic and market forces. In “The Atlas Of Econom-ic Complexity,” for example, the work of Cesar Hidalgo, Ricardo Hausmann and
Network Theory is being applied to un-derstand relationships and interactions in a variety of domains, such as crime prevention and disease management. Dirk Brockmann and Dirk Helbling’s work modeling the spread of infectious diseas-es provides an example of the power that Network Theory holds.
In their article, “The Hidden Geometry of Complex, Network-Driven Contagion Phenomena,” in Science magazine, the authors wrote:
“The global spread of epidemics, ru-mors, opinions, and innovations are com-plex, network-driven dynamic processes. The combined multiscale nature and in-trinsic heterogeneity of the underlying networks make it difficult to develop an
w w w. i n f o r m s . o r g48 | a n a ly t i c s - m aga z i n e . o r g
Data visualizatiOn
project members illustrates how under-standing the composition of a country’s cumulative industrial knowledge can ex-plain economic development in ways not possible through more traditional linear econometric frameworks.
The vast majority of data visualiza-tions today are two-dimensional. How-ever, that’s changing with creative use of color and size, combination of space and time, and advanced computer graphics. For instance, neuroscientists Emmanuelle
Source: The Observatory of Economic Complexity; http://atlas.media.mit.edu/atlas/
j a n U a r y / f e B r U a r y 2 015 | 4 9a n a l y t i c s
the internet of thingsTens of billions of devices will be connected to the Internet in the next decade.
From smart appliances and wearables to automobile sensors and environmental monitors, the Internet of Things will provide unprecedented insight into what’s happening around us. High-throughput, interconnected data streams will help us improve safety, drive operational efficiencies and better understand consumer demand.
In the words of Kevin Ashton, who first coined the term “the Internet of Things” in his seminal 2009 RFID Journal article, “The Internet of Things has the potential to change the world, just as the Internet did. Maybe even more so.”
complexity theoryComplexity Theory posits that many systems are characterized by complex,
non-linear interactions that evolve dynamically and often unpredictably. Known as the “butterfly effect,” small perturbations in one state (“here”) can result in large repercussions in a seemingly unrelated state (“there”). According to Complexity Theory, it’s impossible to predict with certainty a future state, but it is possible to understand the structure and potential states of complex systems.
network theoryNetwork Theory builds on Graph Theory, which applies algorithms to un-
derstand and model pair-wise relationships between objects. Network Theory examines relationship symmetry, with the existence of asymmetric relation-ships providing grounds to predict the likely spread of information (social net-work analysis), dissect complex disorders (biological network analysis), find the shortest path between two points (network optimization) and identify target objects based on their behavior (link analysis).
w w w. i n f o r m s . o r g5 0 | a n a ly t i c s - m aga z i n e . o r g
Data visualizatiOn
dynamic images above or below a plane for more natural exploration and manipu-lation of complex data. And commentary from team members Curtis Wong and David Brown posted on Microsoft News suggests that Holograph may one day al-low users to actually reach inside a visual and interact with it.
Tognoli and Scott Kelso developed a five-dimensional model known as the 5-D colorimetric technique, that provides a dynamic and comprehensive view of brain activity through spatiotemporal dis-play and color coding. Another example is Microsoft’s Holograph, an interactive 3-D platform that can render static and
Source: Florida Atlantic University, Center for Complex Systems and Brain Sciences; http://www.ccs.fau.edu/hbbl3/?p=1013
Multidimensional visualizationThe adage “a picture is worth a thousand words” gained credence from our
ability to process visuals more easily than text. Visualization has also been shown to improve learning and recall, and can portray complex concepts and relation-ships more easily than can text. Recent developments in computer graphics are making possible visualizations that enable the integration, manipulation and ex-ploration of dynamic multidimensional data sets. Multidimensional visualizations allow users to not only examine data from new perspectives but also interact with it more effectively.
j a n U a r y / f e B r U a r y 2 015 | 51a n a l y t i c s
As the world becomes increasingly in-terconnected and interdependent, oppor-tunities to generate value through data visualization will only increase. The Inter-net of Things will have a profound effect on the role that data visualization can play in organizations and society, improving our ability to understand how humans and machines interact with each other and the environment. Application of evolving cog-nitive frameworks, such as Network and
Complexity Theories, will help us better reflect dynamic and intricate structural dependencies. And advances in multidi-mensional visualization will allow us to more effectively synthesize and explore spatiotemporal conditions. ❙
Will Towler ([email protected]) is an analytics and insights specialist. The views expressed in this article are those of the author and do not necessarily represent the views of an employer or business partners.
figure 7: Microsoft’s Holograph, an interactive 3-D platform.
Request a no-obligation infORms member benefits packetFor more information, visit: http://www.informs.org/Membership
w w w. i n f o r m s . o r g52 | a n a ly t i c s - m aga z i n e . o r g
he 1977 film “Powers of Ten” portrays the universe as an arena of both conti-nuity and change. The short
documentary, selected by the Library of Congress as being “culturally, histori-cally or aesthetically significant,” and written and directed by Charles and Ray Eames, begins with a 1-meter distant shot of a man laying by a picnic setting and steadily moves out until it reveals the very edge of the known universe. Then, at a rate of 10-to-the-tenth meters per
second, the film rushes us back toward Earth to the reclined man’s hand and fur-ther down to the level of a carbon atom on his skin. That fascinating journey into the macro and then micro demonstrates visually the importance of scale and, in a metaphysical sense, the importance of visualization itself.
The importance of data visualization becomes more obvious when viewed within the context of how the human brain works. Much has been written in recent years about how the processes of the
j a n U a r y / f e B r U a r y 2 015 | 53a n a l y t i c s
brain and how understanding those pro-cesses can provide profound insights. In his best-selling 2012 book “Thinking, Fast and Slow,” Nobel laureate Daniel Kahneman introduces the terms System 1 and System 2. The terms differentiate between the information processing that occurs in the human subconscious and conscious minds. System 1 addresses the functions that are uncontrolled and effortless. System 2 comprises func-tions that are controlled and require effort to engage. In action, System 1 allows us to instantaneously recognize facial expressions – visual processing. In contrast, System 2 allows us to make complex decisions or apply reason.
A little more than a decade before the release of Kahneman’s book, Danish
physicist Tor Nørretranders, in his book “The User Illusion: Cutting Consciousness Down to Size,” converts the “bandwidth of human senses” to computer terms. He explains just why data visualization (a manifestation of System 1) is perhaps the most powerful form of data interpretation. Nørretranders demonstrates that when assessing the “language of the mind,” the sense of sight simply operates at an order of magnitude faster than the sense of touch (similar to the bandwidth associ-ated with a network of computers), which in itself operates at an order of magnitude faster than the sense of smell. As such, the sense of smell operates at an order of magnitude faster than the sense of taste (which has a bandwidth similar to a calculator)!
Source: “The User Illusion: Cutting Consciousness Down to Size,” by Tor Nørretranders (Penguin Press Science)
figure 1: Natural log of relative operating speeds.
w w w. i n f o r m s . o r g54 | a n a ly t i c s - m aga z i n e . o r g
Data visualizatiOn
Realizing how quickly we understand and in-ternalize what we see is at the foundation of what makes data visualization such an important as-pect of how we analyze information and make better decisions. That said, the mechanisms be-hind data visualization create a powerful tool to design effective visualizations to suit any context – whether that tool is a simple, static bar chart or something vastly more complex, multidimensional and interactive. As such, the science behind data visualization ranges from the fundamentals of how we literally see to the complexities of cognitive psychology.
Combining the science with the art – how best to portray the intent of any particular visualization – winds up somewhere on a curve between pre-sentation and exploration. The difference between presentation and exploration can be described as the difference between presenting a known story in a data set using analysis and exploring a not-yet-understood data set using a visual examina-tion. Henry David Thoreau said, “It’s not what you look at that matters, it’s what you see.” With data visualization, the significance of the quote is quite literal. It’s a fully formed discipline that requires multiple skills – among them, the knowledge of statistics, ideas of space, design and topography, and a deep subject matter expertise in the sector being served.
FRom thEoRy to PRaCtiCE
Currently, for any company that deals with a titanic amount of data, data visualization is and will remain an absolutely fundamental tool. Verisk
Realizing how quickly we understand and internalize
what we see is at the foundation of what makes data visualization such an
j a n U a r y / f e B r U a r y 2 015 | 55a n a l y t i c s
promotional campaign at its retail stores through the holiday season. The man-agement team at the retailer would in-variably want to know: Do we know the baseline sales at our stores and our com-petitor stores before the promotional pe-riod? Let’s say maybe. Do we know how shoppers at our stores respond to promo-tional offers in the regular season? That’s another maybe. Do we know what the weather was like and if it played a role in affecting shopper turnout at our stores during the campaign period? That’s one more maybe. Would all those pieces of information come together at the same time and be presented to management in a manner that’s easy to analyze? That’s highly unlikely. And that’s a great exam-ple of where visualization becomes so helpful.
As consumers of information, we’re all demanding visualization in our own way. We’ve started to reject the culture of sound bites and nonsynthesized statistics that agenda-driven interest groups have inundated us with in the last two decades. Visualization allows us to map the infor-mation in a way that leads to better deci-sion-making – easier and faster. The 2012 InformationWeek Business Intelligence, Analytics and Information Management Survey, conducted in late 2011, indicated nearly half (45 percent) of the 414 respon-dents cited “ease-of-use challenges with
Analytics is a prime example – collecting and maintaining highly granular data on several billions of insurance policies and claims, credit card and debit card trans-actions, real estate, health services, gov-ernment and human resources. While many consumer-centric firms have long been skilled in collecting information, they now generate and acquire exponentially growing, disparate and complex quanti-ties of data – and depend on that data in many ways for their very survival in to-day’s marketplace.
Much of the talk today about data management and analysis and its effect on how business gets done targets the science of analytical modeling. In that pursuit, many firms indeed have come far, and yet they still have farther to go. Large, well-capitalized firms (such as banks, in-surers, retailers) spend considerable re-sources and energy collecting and storing data, not just because they produce a lot of it but more likely because the regulatory environment mandates storing much of it. These firms haven’t spent nearly enough effort aggregating their data across func-tional silos, integrating internal data with third-party data, analyzing the data, and distributing the resulting insights to peo-ple who can take action on it.
As an example from the retail sec-tor, imagine that a retailer is looking to assess the effectiveness of a particular
w w w. i n f o r m s . o r g5 6 | a n a ly t i c s - m aga z i n e . o r g
Data visualizatiOn
complex software/less-technically savvy employees” as the second-biggest barrier to adopting business intelligence/analytics products – fractionally behind the biggest barrier, “data quality problems,” cited by 46 percent of respondents.
mothER natuRE’s inFogRaPhiCs
Catastrophe risk management has come a long way in its 25-year history, and the sophistication of those analytics goes well beyond the numbers in a database. The end result has been fast, intuitive
insight into what drives risk.Looking back to Superstorm Sandy,
healthcare officials in New York City, in ad-vance of the storm, were trying to decide whether to evacuate hospitals. In the end, many chose not to move patients before the storm. Unfortunately, numerous hos-pitals were then catastrophically flooded, and patients had to be moved during the worst of the deluge.
Certainly, myriad factors go into as-sessing a situation like that, but as ana-lytics and their visualization become
Source: AIR Touchstone® zoomed-in surge
figure 2: A detailed inundation map of the New England coastline depicted in Figure 2 shows the surge footprint of Hurricane Sandy in blue. The orange circles represent clustered locations by actual number, with the largest circles containing the most indi-vidual locations. Where the blue and orange overlap, the map illustrates where Sandy had the greatest impact. Such maps help companies determine the extent of floods and resulting losses.
j a n U a r y / f e B r U a r y 2 015 | 5 7a n a l y t i c s
increasingly sophisticated, they’ll be able to help risk-bearing organizations, includ-ing insurers and local authorities, develop appropriate prescriptions for mitigating risk – by providing the contextual detail for better-informed decisions.
Today’s advanced climate models are capable of effectively projecting the impact of storms as they get closer to coastlines or geographic regions. Such models can assess the total number of homeowners expected to be affected, when an event is expected to worsen, and when it will be safe for insurance personnel to move into the area. The visualization models enable the decision-maker or assessor to evalu-ate locations at the individual building level. That capability facilitates a preplan-ning process and allows companies to communicate proactively with policyhold-ers so they can take certain loss control measures – such as boarding windows, reducing chance of fire, and so on – to mitigate damage. Such models are also allowing insurers to readily project and vi-sualize the impact of fallen trees on pow-er lines serving a group of policyholders.
Given the complexity associated with climate change and the inherent difficulty in the assimilation of evolving and inter-dependent data, our dependence on a sophisticated and constantly improving visualization capability is far too great to be denied.
thE sight in BusinEss insight
Unquestionably, the tried-and-true bar, line and pie charts have served us well. But when the complexities of re-lationships are more nuanced and the data becomes more unstructured, visual analytics need to become more dynamic, multidimensional and customized.
For lenders and insurers, visualiza-tion can help identify a range of data issues quickly – from a high-level view of exposure location to exposure com-position and completeness, including breakdowns by profile of the entities at risk (customers, businesses, prop-erties, vehicles and so on). Visual link analysis technology helps discover critical, previously hidden connections within data. Seeing those connections – within proprietary data, in data from external sources or through a combi-nation of sources – provides insight and knowledge to make decisions. The technology finds all data elements ap-plicable to a question and draws a pic-ture of the connections among those elements, revealing previously invisible relationships. The contextual approach provides a multidimensional under-standing of profitability, customer be-havior, and industry trends.
Data integrity can be a significant problem for large organizations, especial-ly where multiple, complex databases are
w w w. i n f o r m s . o r g58 | a n a ly t i c s - m aga z i n e . o r g
Data visualizatiOn
involved. Mapping techniques often find thousands of errors in a fraction of the nor-mal time. Mapping also finds red flags in claims data. Fraud investigators at finan-cial institutions often use visual link analy-sis to assist in their inquiries. For example, a money-laundering investigator monitors each check, credit card or ATM withdrawal over a specific threshold, and the technol-ogy helps in instantly flagging irregular pat-terns, revealing potential sources of fraud or money laundering. Seeing those con-nections – within company data, in data from external sources or through a combi-nation of sources can give claims investi-gators insight and knowledge to help make better decisions.
Visualization is useful in insurance for commercial fleet and personal auto poli-cyholders. Telematics programs use sen-sors to determine factors as simple as distance (vehicle miles traveled) and as sophisticated as camera-based record-ing. Devices transmit and store the resul-tant collection for immediate or deferred analysis, meaningful interpretation and visualization. Although the use of telemat-ics data and visualization is in its early
stages, the usage-based insurance (UBI) opt-in rate is expected to increase to 20 percent over the next five years, accord-ing to one recent industry poll. Other polls consistently show that two-thirds of con-sumers are open to telematics-based in-surance policies, especially if there’s the potential for premium discounts. Among newer consumers of vehicle insurance — the Gen Ys and the Millennials — the use of telematics and visualization technology is almost expected.
While throughout history and in the present day there is always that rare breed with the unique and innate ability to quickly make sense of disparate sources of infor-mation and data, the mortals among us are blessed to be living at a time replete with the data and tools to make those con-nections for us in a fraction – enabling us not only to make better business decisions but maybe even allow us to see the as yet unforeseen. ❙
Dr. Nana Banerjee is a group executive of Verisk Analytics. He serves as president of Argus Information and Advisory Services and as chief analytics officer of Verisk Analytics.
join the analytics section of infORmsFor more information, visit: http://www.informs.org/Community/Analytics/Membership
▪ Continuing Ed courses for Analytics Professionals
▪ Online access to the latest in operations research and advanced analytics techniques
▪ Unsurpassed Networking Opportunities available in INFORMS Communities and at Meetings
▪ Subscriptions to online and print INFORMS Publications
▪ INFORMS Career Center - the industry’s leading job board
JOIN
Is the largest association for analytics in the center of your professional network?
It should be.
w w w. i n f o r m s . o r g6 0 | a n a ly t i c s - m aga z i n e . o r g
he outbreak of Ebola disease in West Africa, with some cases spreading elsewhere, has been a source of con-
cern and frustration throughout the world. This is the story about how a small data science and analytics company found a way to help contain the spread of the virus.
In September 2014, Brian Umana, a consultant at Illumina Consulting Group
(ICG), had a conversation with compa-ny CEO David Waldrop about the Ebola epidemic. They each expressed concern about the scale of the human suffering, but also about facets of security and pub-lic health that could limit the spread of the epidemic. Given the nature of their earlier work in real-time streaming data analytics, they eventually began speaking about various ways one might apply data
j a n U a r y / f e B r U a r y 2 015 | 61a n a l y t i c s
to Area 2 (farther from the coast) and then exited Area 2 to create a map-based alert of the boat’s location, satellite data, name, flag, stated destinations and other information. He then implemented more sophisticated versions of this model. The result was additional notice to ports, and to authorities outside of the specific ports in question, about boats that should re-ceive specific attention. Waldrop’s exper-iment worked, and he promptly donated the results to a federal agency.
Owing to the highly contagious nature of Ebola and the close quarters onboard ships, there’s a high risk of an outbreak rapidly affecting an entire crew. Such a ship entering port would pose a grave health threat to persons in the port and surrounding area. U.S. authorities and shippers are aware of the Ebola Maritime Vector threat, but detection and preven-tion measures appear less rigorous than those in the air travel sector. The rules in effect for maritime traffic were and are that ships must file a notice of arrival 96 hours prior to U.S. landfall, and those no-tices must include declaring visits made to any Ebola risk ports during the ship’s last five port visits. In addition, ship captains or owners are required to self-report any communicable disease onboard during the 15-day period prior to entering port.
If such a report is filed, the appropriate U.S. authorities are notified and the ship
science and analytics to the problem. They were not the first to do this, nor in any way did they take the most dramatic or significant action. However, within days of the conversation, ICG began to make a small contribution to safety in the U.S. response to the “maritime vector” of the Ebola virus.
Waldrop built a new-use case for LUX software, ICG’s principal product, a high-volume data streaming and pat-tern recognition tool. What Waldrop had noticed was that detection and preven-tion measures for potential maritime- and port-based spread of the Ebola virus ap-peared less rigorous than corresponding detection and prevention measures used in air travel. No one would dispute that it is essential to monitor airports where passengers are arriving from Ebola-stricken countries as a measure for con-taining the spread of the disease. West African seaports and merchant shipping, however, are another potential means for spreading the disease abroad, a topic that has drawn far less attention.
Waldrop used the software to draw an area of interest off the coast of West Africa, enclosed by a second area of in-terest encompassing all the water out-side of the first area but still inside of a specific radius of the West African coast. Waldrop instructed LUX that any time a boat crossed from Area 1 (the port area)
w w w. i n f o r m s . o r g62 | a n a ly t i c s - m aga z i n e . o r g
tRacking ebOla
may be boarded, inspected and possibly quar-antined. This scenario created compelling incen-tives for non-compliance with disease reporting, as ship owners face major commercial revenue losses if their ships are quarantined, or if contract conditions are not met owing to delays, or if the ship is diverted and fails to reach the port speci-fied in a bill of lading. Taking this situation into account, Waldrop’s experiment with LUX helped address some issues of public safety in ports.
Two federal agencies looked at the infor-mation. It brought the maritime aspect of the problem to their attention and demonstrated a relatively easy way to pay attention to it. They asked about the data sources and began mon-itoring them. The ships were monitored on the LUX watch dashboard. This procedure provided several days of advance warning about any ship that had visited West Africa and was headed to a U.S. port.
EBola maRitimE VECtoR thREat
To mitigate the Ebola threat, LUX uses two real-time data streams to perform analysis. The first is data streamed from internationally man-dated [1] automatic identification system (AIS) reporting. AIS, installed on all merchant ships in the world, is especially relevant to monitoring ships traveling to and from West Africa ports. The second source is data generated by the Global Data on Events, Location and Tone (GDELT) project [2].
The LUX Ebola Maritime Vector monitoring methodology works as follows:
j a n U a r y / f e B r U a r y 2 015 | 63a n a l y t i c s
• A user defines geographic areas of interest (AOI). In this case, two such ar-eas are displayed on the map as shown in Figure 1.
• The user then writes rules instruct-ing the software to generate alerts on any inbound or outbound ship crossing the boundaries of the AOIs.
• When an alert is first received on an inbound or outbound ship, AIS data is used to determine the ship’s name, its AIS unique identifier number and other data such as course, speed, flag, de-clared destination port and cargo embed-ded in the AIS reporting.
• Once the ship is identified, the user
writes another rule containing the ship’s name and AIS identifier instructing the software to track and generate alerts on its location and activity.
• LUX tracks the ship to its next port (declared or not) and sends the user alerts based on rules and geographic ar-eas of interest related to that port. A use-ful tool in this regard is the dynamic area of interest (DAOI). The DAOI is centered on the ship itself and moves with it. A DAOI of any radius may be established. For example, an alert could be generated anytime a ship is within 50 nautical miles of land, providing warning of a pending port visit.
figure 1: Illustration of the LUX user interface. The outer AOI provides alerts on arrivals and departures from the zone of concern. The inner AOI monitors activity in and around port areas.
w w w. i n f o r m s . o r g6 4 | a n a ly t i c s - m aga z i n e . o r g
tRacking ebOla
• The user then instructs the software to generate alerts on information col-lected by GDELT that matches the user’s rule parameters such as reports of Ebola cases, quarantining of ships, illness among ship crew members, etc. This is a potential source for discovering Ebola’s spread through the maritime vector.
• LUX also provides the means to de-tect abnormal behavior, such as a ship deviating from course, diverting to a port other than its declared port destination, or, particularly, cessation of its AIS re-porting stream.
• If a ship’s AIS reporting were to cease, the software’s forecasting analyt-ics would still provide an estimated track and the ship’s progress along it.
• To facilitate monitoring of multiple ships (thousands), rules alerts can be sent to a user-established watch board. The watch board displays, as color-cod-ed cells, aggregates of AIS alert reporting on as many individual ships as desired. It also monitors and displays GDELT alerts. In this way a user may be relieved of constantly monitoring activity for situ-ational awareness and let LUX take up that task 24/7. When the number or type of alerts reaches a user specified thresh-old, the watch board changes the color of the appropriate watch item cells as a vi-sual notification. An audible tone may be added as an additional notification aid.
What this story highlights is the vast amount of information available at the fingertips of everyone in the analytics community, the ability to act quickly on that information when it is properly ap-plied and brought to the right people’s at-tention, and the potential benefit of such analysis and action. ❙
Doug Samuelson ([email protected]) is president and chief scientist of InfoLogix, Inc., in Annandale, Va., and a senior operations research analyst with Group W, Inc., in Merrifield and Triangle, Va., supporting the Marine Corps Combat Development Command (MCCDC). He is a longtime member of INFORMS.
Brian Umana ([email protected]), a consultant at Illumina Consulting Group, has worked in software analytics, political analytics and media.
RefeRences
1. international Maritime organization’s international convention for the safety of life at sea
2. gdelt is the largest, most comprehensive and highest resolution open source database of human society ever created. gdelt monitors and analyzes the world’s news media from nearly every corner of every country in print, broadcast and web formats, in over 100 languages, every moment of every day.
subscribe to AnalyticsIt’s fast, it’s easy and it’s FREE! Just visit: http://analytics.informs.org/
INFORMS third conference on the health sector brings together researchers and stakeholders around the most current work in healthcare operations research, systems engineering, and analytics in one highly-focused conference.
Cross-cultural view of healthcare systems and analysis of operational impacts.
Structured networking opportunities, including birds-of-a-feather discussion groups and facilitated networking over lunch.
Collegial, small-scale setting in a vibrant location in one of the fastest growing cities in the U.S.
HEALTHCARE 2015NASHVILLE TENNESSEE
Reinventing
Value
$
JULY 29-31OMNI HOTEL
http://meetings.informs.org/healthcare2015
GENERAL CHAIRSRobert Dittus, Vanderbilt UniversityM. Eric Johnson, Vanderbilt University
PROGRAM CHAIRVikram Tiwari, Vanderbilt University
INVITED SESSIONS CHAIREva Lee, Georgia Institute of Technology
SPONSORED SESSIONS CHAIRMehmet Begen, Western University, Canada
HAS STUDENT PAPER COMPETITION CHAIRSAnita Tucker, Brandeis UniversityBaris Ata, University of Chicago
ADVISORY COMMITTEEDr. Warren Sandberg, Vanderbilt UniversityR. Lawrence Van Horn, Vanderbilt University
Abstract Submissions Now Open!http://meetings.informs.org/healthcare2015/abstract-submission
INFORMS Healthcare 2015 invites submissions on research andapplications in the healthcare arena. We welcome submissions in either oral or poster formats, which will be reviewed by the Committee. There is a limit of one oral presentation per presenting/lead author. Authors may present a poster as well as one oral presentation.
w w w. i n f o r m s . o r g6 6 | a n a ly t i c s - m aga z i n e . o r g
INFORMS will return to Huntington Beach, Calif., for the 2015 INFORMS Conference on Business Analytics and Operations Research, which will be held April 12-14 at the Hyatt Regency Resort & Spa. The conference, with the breadth of its coverage and the depth and quality of its presentations, should be of great interest to practitioners of analytics and oper-ations research, whether they work in the public sec-tor or private industry. It will also appeal to academics who are training future analytics practitioners or have research interests that would benefit from exposure to cutting-edge work in industry and government.
The conference will run from Sunday to Tuesday, with the main conference program held on Monday and Tuesday. The Sunday sessions focus on vendor technology workshops, the INFORMS Professional colloquium and the UPS George D. Smith finalist pre-sentations. Plenaries from industry luminaries and thought leaders [such as keynote speaker William Ruh, VP of GE Software; see accompanying sidebar] headline the main conference program.
The conference will celebrate the best in the pro-fession, headlined by the prestigious Franz Edelman Award competition and the Franz Edelman Gala, both
Business Analytics and O.R. Conference set for Huntington Beach
j a n U a r y / f e B r U a r y 2 015 | 67a n a l y t i c s
of which will be held on Monday. There will also be an invited session over two days that will feature presentations by finalists and winners of various other INFORMS prizes such as the Daniel H. Wagner Prize, the INFORMS Prize, the Analytics Innovation Prize and UPS George D. Smith Prize.
The main program will cover a wide range of topics in analytics and opera-tions research practice with tracks on marketing analytics, unstructured data analytics, real-time decision systems, supply chain and logistics, revenue man-agement and pricing, managing risk and decision analysis. Attendees typically have a choice of at least eight parallel,
high-quality presentations to choose from among these areas and can tailor their conference itinerary to their profes-sional interests.
Additionally, software tutorials from vendor sponsors run parallel to the main sessions, allowing attendees an oppor-tunity to hone their skills in that arena. Poster sessions on both days offer further possibilities for attendees to present their work to their peers in an informal setting.
Conference receptions, meals, breaks and other events are designed to offer plenty of networking opportunities.
In summary, analytics and O.R. pro-fessionals who attend this conference will hear about the latest and greatest
Huntington Beach, Calif., will provide the backdrop for
the INFORMS Conference on Business Analytics and
Operations Research in April.
w w w. i n f o r m s . o r g68 | a n a ly t i c s - m aga z i n e . o r g
cOnfeRence pReview
in practice, learn new ideas and best practices, expand their professional and personal networks and celebrate their profession. Local area students are par-ticularly encouraged to attend and take advantage of the reduced conference registration and professional exposure afforded by this conference.
Look for announcements and updates on the INFORMS website and the confer-ence website for up to date information about the conference. ❙
Manoj Chari is chair of the 2015 INFORMS Conference on Business Analytics and Operations Research.
William Ruh, vice president of GE Software, will deliver the key-note address at the 2015 INFORMS Conference on Business Analytics and Operations Research set for April 12-14 in Huntington Beach, Calif.
Ruh joined GE in 2011 to lead a renewed focus on software and analytics and to drive the global strat-egy, direction, operations and portfolio of software-based services that har-ness streams of customized performance data for reaching new levels of productivity. As part of this charter, GE Software is creating an ecosystem of strategic partnerships and investments, developing a com-mon software platform (PredixTM) for powering industrial Internet offerings,
ge’s Ruh to deliver conference keynoteand establishing a portfolio of shared services including cyber security, data science, user experience, cloud and agile development for building more sustainable and safe products.
Prior to joining GE, Ruh was vice president at Cisco, where he held global responsibility for develop-ing advanced services and solutions. He has more than 30 years of industry experience in enterprise application integration and
object-oriented technology. Ruh is an accomplished author and
a frequent speaker on such topics as emerging business models, cloud com-puting, analytics, mobile computing, ag-ile development, large-scale distributed systems and M2M communications.
w w w. i n f o r m s . o r g70 | a n a ly t i c s - m aga z i n e . o r g
by haRRisOn schRamm, cap
I had the opportunity to play Battleship with two boys in my family. For those who may not be famil-iar, Battleship is a board game played in two stages. First, each player places their fleet of five ships cover-ing a total of 17 squares on a 10x10 grid without their opponent seeing. Ships may be placed along rows or columns but not diagonals. Then, players call out grid squares to “shoot” at, and the other player reports if it is a “hit” or a “miss.” It is presumed that the players are honest. The research question is: How can we tell if another player is being dishonest? (Note: this project was much more fun than I expected it to be!)
The basics: When there are no hits on the board, 17/100 = 17 percent of the spaces are occupied. A first approximation or “naïve” plan would say that the expected number of shots until the first hit is 1/.17 = 5.8. This is slightly inaccurate, because the shots are not independent. A finite population adjustment shows that the expected number of shots is approximately
j a n U a r y / f e B r U a r y 2 015 | 71a n a l y t i c s
5.3. This is a minor – but noticeable – dif-ference in the number of times to first hit. I would begin to seriously suspect (95 per-cent confidence) that my opponent was cheating if I did not score a hit in the first 15 attempts, and would be almost certain (99 percent confidence) he was cheating if I did not score a hit in the first 22 shots.
Can we do better? Yes! Each ship is rectangular, and can only inhabit squares that are along ranks or files (like rooks in chess). The way to handle this is to pick shots along diagonals (like bishops
in chess). If we have already shot and missed in a particular square, say, A1, then we don’t have to shoot at A2 or B1. For example, if we are most interested in hitting ships three or more lengths in size – submarine, cruiser, battleship and car-rier – then we may “skip two” and “cover” the board in 33 shots (see Figure 1).
We can think of this as superimpos-ing a couple of different games on top of each other; for example, in the carrier case (skip four), it would look something like this:
X X XX X X
X X X XX X X
X X XX X X X
X X XX X X
X X X XX X X
figure 1: A scheme for covering the Battleship board in 33 shots. There are similar schemes for covering the board in 50 and 20 shots.
figure 2: Picture of the “superimposed” Battleship game. Conceptually, the carrier is occupying one square and is playing on the 20-square board (left), while the rest of the fleet are occupying their respective sizes on the 100 square board. Each shot removes one square off the respective boards.
w w w. i n f o r m s . o r g72 | a n a ly t i c s - m aga z i n e . o r g
five-minute analyst
The general method for determining the probability of hit in the next round, given no hits in the preceding rounds, is to determine the probability of hit for each ship type, and the key intuition is that for ships that are three or larger, the scheme shown in Figure 1 effectively “covers” three spaces on the board after the first shot, which always covers just one (why?). The general formula is:
The formulae for the other spacing are
similar. We can compare these approaches,
and see that they all perform remarkably similar; the five-space approach is slightly better for achieving the first hit.
figure 3: Performance of various skip-ping schemes for scoring the first “hit” in Battleship.
We have only considered the time until the first hit. While beyond the bounds of this analysis, my sense is that “skip two,” as depicted in Figure 1, is the optimum game playing strategy. This strategy has a
possibility of missing the two-square “cor-vette” ship, but will certainly hit all the oth-ers. This is an area ripe for further analysis.
defending along the rows: What does it mean to be “maximally random?” There are different interpretations, but one way might be to make the groupings along the rows and columns of the board fit a Poisson distribution with λ=1.7. A perfect fit is not possible because at least one of the rows or columns will have a “5” due to the carrier, which only has a .3 prob-ability of occurrence. This turns out to be an optimization problem that exceeds the bounds of the Five-Minute Analyst. This problem is very similar to solving Sudoku with optimization. Below is a candidate solution and its performance measured by the Poisson metric.
In conclusion, this piece has brought up more areas for analysis on this topic than presented answers. Hopefully you will think differently about Battleship next time you play.
Bonus: If you don’t have your com-puter handy while you are figuring out which strategy has the greatest rate of improvement, you can appeal to calculus. The proposed strategies are of the form
j a n U a r y / f e B r U a r y 2 015 | 73a n a l y t i c s
1 11 1 1 1 1
111 1
11 1 1 1
1
figure 4: A possible laydown of Battleship. This was developed using a spreadsheet to sum the rows and columns and generate the histogram shown in Figure 5.
figure 5: Performance of the laydown shown in Figure 4. The ideal distribution
is shown in gray, with the row and column distributions shown in blue
and orange, respectively.
Harrison Schramm ([email protected]) is an operations research professional in the Washington, D.C., area. He is a member of INFORMS and a Certified Analytics Professional (CAP).
w w w. i n f o r m s . o r g74 | a n a ly t i c s - m aga z i n e . o r g
thinking analytically
by jOhn tOczek
John Toczek is the senior director of Decision Support and Analytics for ARAMARK Corporation in the Global
Operational Excellence group. He earned a bachelor of science degree
in chemical engineering at Drexel University (1996) and a master’s
degree in operations research from Virginia Commonwealth University
(2005). He is a member of INFORMS.
A new city is being built that will include 20 distinct neighborhoods as shown by the house icons in the map. As part of the planning process, electricity needs to be connected to each of the neighborhoods.
The city has been allocated funds to put in three electrical substations to service the elec-trical needs of the neighborhoods. The sub-stations are represented by the three electrical box icons to the right of the map. Because lay-ing electrical line to each neighborhood is ex-pensive, the placement of the substations on the map requires careful consideration.
A neighborhood will be serviced by the nearest electrical substation. A neighborhood may only be connected to one substation. The substations may be placed in any cell (includ-ing the same cell as an existing neighbor-hood). The cost of electrical wiring is $1 million per kilometer. Distances are measured using a direct line between cells, which are each one kilometer apart. For example, the distance be-tween cell A1 and B2 is 1.41 kilometers.
Question: What is the minimum cost required to connect all neighborhoods to electricity?
Send your answer to [email protected] by March 15. The winner, chosen ran-domly from correct answers, will receive a $25 Amazon.com Gift Card. Past questions can be found at puzzlor.com. ❙
