of 71

Analytics Marchapril 2013

Published on February 2017 | Categories: Documents | Downloads: 18 | Comments: 0
172 views

Comments

Content


H T T P : / / WWW. A N A L Y T I C S - MAGA Z I N E . O R G
ALSO INSIDE:
MARCH/ APRI L 2013 DRIVING BETTER BUSINESS DECISIONS
• Smart Care
The promise of population
health management
• Statistical Software
Survey of products in
data-rich environment
• Data Mining at Dow
Case study: What to make,
when and for whom
RECOMMENDATION
ENGINES REVVED UP
Executive Edge
AYATA CEO
Atanu Basu’s
five pillars of
analytics success
THUMBS UP TO POWERFUL,
WIDELY USED ONLINE
MARKETING TOOLS
BROUGHT TO YOU BY:
WWW. I NF OR MS . OR G 2 | A NA LY T I CS - MAGA Z I NE . OR G
What’s that buzzing sound?
I NSI DE STORY
Five years ago this month, INFORMS
(Institute for Operations Research and
the Management Sciences) published
the frst issue of Analytics magazine.
During the early planning stages, it took
a group of folks on a conference call
(INFORMS offcers and staffers and
Lionheart Publishing personnel) about
one nanosecond to come up with a
name for the publication (Analytics) and
about a minute to come up with a tagline
(“driving better business decisions”).
Anyone who has ever served on an ad
hoc committee knows that has to be
an all-time land speed record for group
decision-making.
Looking back, what amazed me was
not so much the speedy consensus
around the name (what else are you going
to call a magazine about analytics?), but
that the title “Analytics” had not already
been claimed by some other publisher.
After all, a few months before our con-
ference call, the best-selling book “Com-
peting on Analytics” by Tom Davenport
and Jeanne Harris had seemingly and
instantly turned “analytics” into a must-
have for every organization and turned
number-crunching, back-offce analysts
into corporate rock stars. The business
world was all abuzz over analytics, yet
miraculously the magazine title was still out
there, waiting to be claimed by INFORMS,
which grabbed it and ran with it.
Five years later, the business world is still
buzzing over analytics. If anything, the buzz
has grown louder, as information technology
research heavyweights such as Gartner jump
on the bandwagon. But buzz and buzzwords
do not produce a better decision-making pro-
cess, and many organizations and analysts
are still grappling with some basic questions
about analytics, such as: How do we get
started with analytics? How do we achieve
analytics success? Perhaps most important-
ly, does analytics live up to the hype?
Glad you asked. Three of our columnists
address those very questions in this issue of
Analytics. Andy Boyd tackles the frst in his
Proft Center column (“Getting started with
analytics”), Atanu Basu offers insight on the
second in the Executive Edge column (“Five
pillars of prescriptive analytics success”) and
Vijay Mehrotra explores the provocative third
question in his Analyze This! column (“Are an-
alytics and big data overhyped?”).
Five years later, we still have more
questions than answers about analytics,
but we learn something new every day. ❙
– PETER HORNER, EDITOR
peter.horner
@
mail.informs.org
Free Trial and Student Licenses are available at our website,
along with many real life cases of AIMMS' success.
Visit the newly designed AIMMS.com for more information now.
Bellevue, WA, USA Haarlem, the Netherlands Singapore Shanghai, China
+1 425 458 4024 +31 23 5 511 512 +65 6521 2827 +86 21 51160733
www.aimms.com • [email protected]
AIMMS is a registered trademark of Paragon Decision Technology B.V. Other brands and their products are trademarks of their respective holders
AIMMS is the optimization tool of choice for corporations,
educational institutions and consultants worldwide.

Their choice is AIMMS because it is the most flexible mathematical modeling platform available, with:
• an extensive set of solvers
• advanced algorithmic capabilities
• an intuitive user interface builder
• and now AIMMS PRO, the AIMMS app store which allows developers to create a customized
but ready menu of optimization applications for a wide range of users

AIMMS can deliver successful projects of every type because it offers an unrivaled breadth and
depth of mathematical optimization concepts that can be quickly applied to your unique set of
challenges, including all of the following mathematical programming types:
• Linear Programming
• Mixed Integer Programming
• Quadratic Programming
• Mixed Integer Quadratic Programming
• Quadratically Constrained Programming
• Mixed Integer Quadratically Constrained
Program
• Constraint Programming
• Global Optimization
• Nonlinear Programming
• Mixed Complimentary Programming
• Mathematical Program with
Complementarity Constraints
• Mixed Integer Nonlinear Programming
AGILE AGILE
AIMMS
POWERFUL POWERFUL POWERFUL POWERFUL
AGILE AGILE AGILE AGILE OPTIMIZATION OPTIMIZATION
POWERFUL POWERFUL POWERFUL POWERFUL
WWW. I NF OR MS . OR G 4 | A NA LY T I CS - MAGA Z I NE . OR G
DRIVING BETTER BUSINESS DECISIONS
C O N T E N T S
FEATURES
SMART CARE
By Rajib Ghosh, Kylie M. Grenier and Theo Ahadome
The promise of technology-enabled population healthcare
management and its implications.
RECOMMENDATION ENGINES WORK
By Christopher Berry
Barriers to recommendation engines deployment are coming down while
opportunities for deployment are improving.
INTEGRATING DATA MINING AND FORECASTING
By Tim Rey and Chip Wells
Case study: The Dow Chemical Company’s approach to leveraging
time-series data and demand sensing.
STATISTICAL SOFTWARE SURVEY
By James J. Swain
Product comparisons in a data-rich culture driven by intelligent devices,
online commerce and social networks.
CORPORATE PROFILE: BANK OF AMERICA
By Russ Labe
Looking back on 25 years of achievement, the B of A Decision
Support Modeling team has plenty to celebrate.
24
34
42
48
56
48
24
MARCH/ APRI L 2013
Brought to you by
34
Tel 775 831 0300 • Fax 775 831 0314 • [email protected]
ANALYTIC SOLVER PLATFORM
Easy to Use Predictive and Prescriptive Analytics in Excel
How can you get results quickly for business decisions,
without a huge budget for “enterprise analytcs”
sofware, and months of learning tme? Here’s how:
Analytc Solver Platorm does it all in Microsof Excel,
accessing data from PowerPivot and SQL databases.
Sophistcated Data Mining and Predictve Analytcs
Go far beyond other statstcs and forecastng add-ins
for Excel. Use classical multple regression, exponental
smoothing, and ARIMA models, then go further with
regression trees, k-nearest neighbors, and neural
networks for predicton, discriminant analysis, logistc
regression, k-nearest neighbors, classifcaton trees,
naïve Bayes and neural nets for classifcaton, and
associaton rules for afnity (“market basket”) analysis.
Use principal components, k-means clustering, and
hierarchical clustering to simplify and cluster your data.
Simulaton, Optmizaton and Prescriptve Analytcs
Analytc Solver Platorm also includes decision trees,
Monte Carlo simulaton, and powerful conventonal and
stochastc optmizaton for prescriptve analytcs.
Help and Support to Get You Started
Analytc Solver Platorm can help you learn while
getng results in business analytcs, with its Guided
Mode and Constraint Wizard for optmizaton, and
Distributon Wizard for simulaton. You’ll beneft from
User Guides, Help, 30 datasets, 90 sample models, and
new textbooks supportng Analytc Solver Platorm.
Analytc Solver Platorm goes further than any other
sofware with Actve Support that alerts us when you’re
having a problem, and brings live assistance to you right
where you need it – inside Microsof Excel.
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.
6 |
DRIVING BETTER BUSINESS DECISIONS
REGISTER FOR A FREE SUBSCRIPTION:
http://analytics.informs.org
INFORMS BOARD OF DIRECTORS
President Anne G. Robinson, Verizon Wireless
President-Elect Stephen M. Robinson, University of
Wisconsin-Madison
Past President Terry Harrison, Penn State University
Secretary Brian Denton,
University of Michigan
Treasurer Nicholas G. Hall, Ohio State University
Vice President-Meetings William “Bill” Klimack, Chevron
Vice President-Publications Eric Johnson, Dartmouth College
Vice President-
Sections and Societies Paul Messinger, University of Alberta
Vice President-
Information Technology Bjarni Kristjansson, Maximal Software
Vice President-Practice Activities Jack Levis, UPS
Vice President-International Activities Jionghua “Judy” Jin, Univ. of Michigan
Vice President-Membership
and Professional Recognition Ozlem Ergun, Georgia Tech
Vice President-Education Joel Sokol, Georgia Tech
Vice President-Marketing,
Communications and Outreach E. Andrew “Andy” Boyd,
University of Houston
Vice President-Chapters/Fora Olga Raskina, Con-way Freight
INFORMS OFFICES
www.informs.org • Tel: 1-800-4INFORMS

Executive Director Melissa Moore
Meetings Director Teresa V. Cryan
Marketing Director Gary Bennett
Communications Director Barry List

Headquarters INFORMS (Maryland)
7240 Parkway Drive, Suite 300
Hanover, MD 21076 USA
Tel.: 443.757.3500
E-mail: [email protected]
ANALYTICS EDITORIAL AND ADVERTISING
Lionheart Publishing Inc., 506 Roswell Street, Suite 220, Marietta, GA 30060 USA
Tel.: 770.431.0867 • Fax: 770.432.6969
President & Advertising Sales John Llewellyn
[email protected]
Tel.: 770.431.0867, ext.209
Editor Peter R. Horner
[email protected]
Tel.: 770.587.3172
Art Director Lindsay Sport
[email protected]
Tel.: 770.431.0867, ext.223
Advertising Sales Sharon Baker
[email protected]
Tel.: 813.852.9942
Analytics (ISSN 1938-1697) is published six times a year by
the Institute for Operations Research and the Management
Sciences (INFORMS). For a free subscription, register at
http://analytics.informs.org. Address other correspondence to
the editor, Peter Horner, [email protected] The
opinions expressed in Analytics are those of the authors, and
do not necessarily refect the opinions of INFORMS, its offcers,
Lionheart Publishing Inc. or the editorial staff of Analytics.
Analytics copyright ©2013 by the Institute for Operations
Research and the Management Sciences. All rights reserved.
DEPARTMENTS
Inside Story
Executive Edge
Profit Center
Analyze This!
Career Builder
Conference Preview
Five-Minute Analyst
Thinking Analytically
2
8
14
18
22
62
66
70
70
22
Tel 775 831 0300 • Fax 775 831 0314 • [email protected]
ANALYTIC SOLVER PLATFORM
From Solver to Full-Power Business Analytics in Excel
The Excel Solver’s Big Brother Has Everything You
Need for Predictve and Prescriptve Analytcs
From the developers of the Excel Solver, Analytc Solver
Platorm makes the world’s best optmizaton sofware
accessible in Excel. Solve your existng models faster,
scale up to large size, and solve new kinds of problems.
From Linear Programming to Stochastc Optmizaton
Fast linear, quadratc and mixed-integer programming is
just the startng point in Analytc Solver Platorm. Conic,
nonlinear, non-smooth and global optmizaton are just
the next step. Easily incorporate uncertainty and solve
with simulaton optmizaton, stochastc programming,
and robust optmizaton – all at your fngertps.
Ultra-Fast Monte Carlo Simulaton and Decision Trees
Analytc Solver Platorm is also a full-power tool for
Monte Carlo simulaton and decision analysis, with a
Distributon Wizard, 50 distributons, 30 statstcs and
risk measures, and a wide array of charts and graphs.
Comprehensive Forecastng and Data Mining
Analytc Solver Platorm samples data from Excel,
PowerPivot, and SQL databases for forecastng and data
mining, from tme series methods to classifcaton and
regression trees, neural networks and associaton rules.
And you can use visual data exploraton, cluster analysis
and mining on your Monte Carlo simulaton results.
Find Out More, Download Your Free Trial Now
Analytc Solver Platorm comes with Wizards, Help, User
Guides, 90 examples, and unique Actve Support that
brings live assistance to you right inside Microsof Excel.
Visit www.solver.com to learn more, register and
download a free trial – or email or call us today.
WWW. I NF OR MS . OR G 8 | A NA LY T I CS - MAGA Z I NE . OR G
As the Big Data Analytics space continues to
evolve, one of the breakthrough technologies that
businesses will be talking about in the coming years
is prescriptive analytics. The promise of prescriptive
analytics is certainly alluring: it enables decision-mak-
ers to not only look into the future of their mission criti-
cal processes and see the opportunities (and issues)
that are potentially out there, but it also presents the
best course of action to take advantage of that fore-
sight in a timely manner. What should we look for in a
prescriptive analytics solution to ensure it will deliver
business value today and tomorrow?
Consider the following fve pillars to prescriptive
analytics success:
1. HYBRID DATA
Most businesses today run on structured data – num-
bers and categories. According to IBM, 80 percent of the
data currently produced is unstructured – text, image,
video and audio. While some businesses may choose to
run the same way in the future as they do today, doing
so could render them unproductive and noncompetitive.
These businesses may not survive as their customers,
suppliers and competitors move beyond them by taking
Five pillars of prescriptive
analytics success
BY ATANU BASU
While some businesses
may choose to run the
same way in the future
as they do today,
doing so could render
them unproductive and
noncompetitive.
EXECUTI VE EDGE
MAR CH / A P R I L 2013 | 9 A NA L Y T I C S
full advantage of hybrid data, a combination
of unstructured and structured data. Hybrid
data empowers businesses to use all the
available data to make the best decisions
possible.
For a prescriptive analytics technol-
ogy to be transformative, it must be able
to process hybrid data. Without incor-
porating hybrid data, decision-makers
are making their decisions based on just
20 percent of the available data. Figure
1 is a chart from Gartner Research that
showcases the evolution of analytics,
culminating in prescriptive analytics with
hybrid data.
Processing of hybrid data brings into
the mix new technologies that are es-
sential ingredients in prescriptive analyt-
ics software, including computer vision,
speech recognition, image processing,
natural language processing, signal
processing and more. While
traditional disciplines such as
applied statistics remain in-
valuable, they aren’t designed
to process image, video, au-
dio and text.
2. INTEGRATED
PREDICTIONS &
PRESCRIPTIONS
Prescriptive analytics is
about seeing and then shap-
ing the future. Common sense
tells us that one needs to frst see the fu-
ture before one can shape it. The func-
tions – predictions and prescriptions
– must work synergistically for prescrip-
tive analytics to deliver on its promise.
The symbiotic integration of predictions
and prescriptions is the key to wide-
spread adoption and inherent value of
prescriptive analytics.
Assume a scenario where predictions
and prescriptions are coming from two dif-
ferent systems that have been cobbled
together (easier said than done, but let’s ig-
nore that for now). Say this software combo
produced a prescription that turned out to be
faulty. If this is a software issue, is this error
due to a bug in the prediction software or the
prescription software or both? Imagine the
disruption to your business as you investi-
gate the root cause and attempt to preempt
similar erroneous prescriptions in future.
Figure 1: The evolution of analytics according to Gartner.
Source: Gartner Symposium, Orlando, October 2012
WWW. I NF OR MS . OR G 10 | A NA LY T I CS - MAGA Z I NE . OR G
3. PRESCRIPTIONS & SIDE EFFECTS
Prescriptions – i.e., recommended,
time-dependent actions to improve the
future – in prescriptive analytics technol-
ogy are generated using several meth-
ods. A prevalent method of coming up
with prescriptions is through a guided
framework of business rules. This rule
framework can be simple or complex, de-
pending on the business process or the
initiative that is being governed by pre-
scriptive analytics. A more scientifc and
rigorous way to produce prescriptions to
improve the future is through operations
research (O.R.), the science of data-driv-
en decision-making. O.R. takes into ac-
count the objectives, the constraints and
the actionable knobs (known as decision
variables) to produce the best course of
action – a prescription – that doesn’t lead
to undesirable side effects. Both optimi-
zation and simulation technologies, two
prominent branches within O.R., can be
used to generate effective prescriptions.
For a prescriptive analytics technolo-
gy to scale, the solution should use both
business rules and operations research
and use them synergistically. Then, and
only then, this technology will be able to
generate the most effective and timely
prescriptions that the available data will
allow. For the Internet of Everything (or
the Industrial Internet) to reach its true
potential, prescriptive analytics – and the
resulting decision automation – has to
play a pivotal role.
4. ADAPTIVE ALGORITHMS
Image you are driving to work. As you
drive, what you see ahead through your
windshield keeps changing and what
you do based on what you see ahead –
and when you see it ahead – also keeps
changing. In our daily lives, we fnd this
reality to be an obvious one, and we do
the needful without much thought.
Now think about a business process
that you are trying to improve through
prescriptive analytics technology. As this
business process evolves over time, the
technology should continually re-predict
and re-prescribe so the predictions and
prescriptions remain relevant.
In a world of growing data volume, ve-
locity and variety, the prescriptive analytics
technology must be able to automatically
recalibrate all its built-in algorithms, plus
automatically create new algorithms. This
total recalibration also needs to be adap-
tive – dynamic and/or continual – in order
to successfully assist the business process
being managed in an ongoing fashion.
EXECUTI VE EDGE
Subscribe to Analytics
It’s fast, it’s easy and it’s FREE!
Just visit: http://analytics.informs.org/
The Optmizaton / Simulaton Library that Speaks the
Languages of Desktop and Web Computng.
No other “callable library” matches the fexibility of
Solver SDK Platorm – from solving Excel models on
your server, to solving models created in your browser.
Build Your Model Faster in Excel or Visual Studio.
Use Risk Solver Platorm or its subsets to build your
model in Excel. Use Microsof Visual Studio to create
your model in C++, C# or VB.NET. Use C, R, MATLAB,
Java, PHP or JavaScript – it’s your choice.
Use Optmizaton and Simulaton as a Web Service.
Run the Solver Server that comes with Solver SDK on
any PC or server, then solve your models “over the
wire” using our standard Web Service, SOAP and WSDL
protocols. Get real-tme progress, or just disconnect
and retrieve the soluton later – it’s your choice.
Solve Your Most Challenging Linear Mixed-Integer,
Nonlinear and Stochastc Optmizaton Models.
Solver SDK’s wide choice of built-in and plug-in Solver
Engines handle every type and size of optmizaton
problem, from Gurobi and Xpress-MP to Mosek and
KNITRO. Use the SDK’s built-in simulaton optmizaton
and stochastc programming algorithms, and support
for mult-core CPUs.
Download a Free Trial at www.solver.com.
Learn more and download a FREE trial version of Risk
Solver Platorm, Solver SDK Platorm, and our eight
plug-in Solver Engines from Solver.com
THE MOST FLEXIBLE WAY TO DEPLOY YOUR MODEL
SOLVER SDK PLATFORM
Tel 775 831 0300 • Fax 775 831 0314 • [email protected]
WWW. I NF OR MS . OR G 12 | A NA LY T I CS - MAGA Z I NE . OR G
Recalibration of the algorithms in the pre-
scriptive analytics software can be triggered
in several ways – with new data arrival, with
data change(s), after a specifed time pe-
riod and more.
Another important factor to keep in
mind is the “action-ability” of the prescrip-
tions; it is, of course, different for different
business processes. Sometimes, it may
not be helpful to automatically gener-
ate newer prescriptions if the old ones
haven’t been acted upon properly.
5. FEEDBACK MECHANISM
How would the prescriptive analytics
software know if its prescriptions are being
acted upon? Prescriptions are, generally
speaking, time-sensitive action plans in-
volving changes to some actionable infu-
encers to preempt one or more predicted
issues (or to beneft from one or more pre-
dicted opportunities). If a business manag-
er decides to ignore a prescription from the
software, this inaction would at some point
be refected in the incoming data that is
being collected on the actionable infuenc-
ers. The consequence of inaction, if any,
will then be subsequently refected in the
upcoming predictions and prescriptions.
For example, due to lost time, inaction on
a valid prescription could lead to additional
expenses in preempting an upcoming is-
sue that has been fagged (via prediction)
and addressed (via prescription) in the last
round.
While this may change in the near
future, today, there is a difference be-
tween prescriptive analytics software
and prescriptive automation. Prescriptive
automation (Google Car, for example)
has elaborate, built-in process control
(software, hardware, frmware and much
more) to automatically “action” the pre-
scriptions coming from the software side.
Prescriptive analytics software today still
requires human assistance to carry out
these prescriptions.
While Google can probably outft a
car anyway it wants to, companies with
highly sophisticated prescriptive analyt-
ics software are dependent on humans
to act on the prescriptions coming out of
the software. Envision a future where this
distinction would disappear and prescrip-
tive analytics software will become a fully
integrated and embedded component of
the business process it is improving –
and not the new thing that it is today. ❙
Atanu Basu ([email protected]) is the CEO
and president of AYATA, a prescriptive analytics
software company based in Austin, Texas. AYATA’s
customers include Apache Corporation, Cisco
Systems, Microsoft and Dell.
EXECUTI VE EDGE
Help Promote Analytics
It’s fast and it’s easy! Visit:
http://analytics.informs.org/button.html
A membership in INFORMS will help!
visit http://join.informs.org to join online
▪ New in 2013! Certification for Analytics Professionals
▪ New in 2013! A FREE Community Membership
▪ 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
▪ Education programs around the world to enhance your
professional development and growth!
How will you stand out from the crowd?
2013
INFORMS Renewals
available online
Renewing online is quick
and easy. Renew online
to help us stay green.
http://renew.informs.org
WWW. I NF OR MS . OR G 14 | A NA LY T I CS - MAGA Z I NE . OR G
It can be hard breaking into a new area. Very
early in my career I worked at a national labora-
tory developing real-time communication systems
for the U.S. military. In college I’d taken a class in
FORTRAN, but that was the extent of my experi-
ence with computers. In my new position I had to
learn the ins and outs of military communication
protocols, how computer hardware worked (from
antennas to the latest microprocessors), and how
to program in a new language called C. I was im-
mersed in a world entirely foreign to me. But with
a little experience, working side by side with capa-
ble, supportive people who mentored me, I came
up to speed surprisingly quickly. Looking back, it
was one of the most enjoyable times of my life be-
cause I was learning and doing new things every
day.
Analytics can be like that, too. But for many peo-
ple the real challenge is how to get started – what’s
the frst step?
INFORMS – the non-proft professional society re-
sponsible for publishing this magazine – is trying to
make that easier (for more on INFORMS, see “What
is INFORMS and Why Should I Care?”). Why? Be-
cause it’s part of the society’s charter. Promoting the
growth of analytics and operations research is one
way INFORMS serves its membership.
Getting started with
analytics
BY E. ANDREW BOYD
For many people the
real challenge is how to
get started – what’s the
first step?
PROFI T CENTER
MAR CH / A P R I L 2013 | 15 A NA L Y T I C S
INFORMS recently released a new
website – actually, a new collection of
pages linked to the society’s main Web
portal – devoted specifcally to helping
people get started with analytics. To ac-
cess it, visit http://informs.org/About-IN-
FORMS/What-is-Analytics (which itself
has good information about analytics)
and click on the link Getting Started With
Analytics.
The case for using analytics is an im-
portant theme once landing on the web-
site. Podcasts and videos are available,
as are testimonials and success stories.
Success stories can be chosen by indus-
try, function or beneft, and they provide
a look at how others have used analytics
to their advantage.
One story, entitled “Coca-Cola Enter-
prises: Optimizing Product Delivery of 42
Billion Soft Drinks a Year,” describes how
the company reworked the scheduling
of its delivery trucks, saving tens of mil-
lions of dollars a year. Another describes
how the Industrial and Commercial Bank
of China upgraded and expanded its
THE DANIEL H. WAGNER PRIZE
Excellence in Operations Researc Practice
Apply to win this prestigious practice prize that rewards professionals who
devise innovative analytical methods, utilize those methods is a verifiably
successful O.R./analytics project, and describe their work in a clear,
well-writen paper.

Two-page abstract is due by May 1, 2013.
Visit www.informs.org/wagnerprize for details.
This top INFORMS practice prize spans all O.R. and analytics disciplines
and application fields. Any work presented in an INFORMS section or society
practice-oriented competition is eligible as long as the work did not result in
a published paper.

Daniel H. Wagner
The Wagner Prize competition is high-profile, with its own track at INFORMS Annual Meeting. Presentations
are widely distributed via streaming video. Finalist papers are published as a special issue in INFORMS
respected practice journal Interfaces.

The 2013 competition will be held at the INFORMS Annual Meeting, October 6-9, in Minneapolis, Minnesota.
First-place prize of $1,000 will be awarded at the Edelman Gala, during the April 2014 Conference on Business
Analytics and O.R. in Boston, Massachusets.
2013
View the 2012 Wagner Presentations online at the INFORMS Video Learning Center:
htp://www.livewebcast.net/INFORMS_AM_Wagner_Prize_2012
WWW. I NF OR MS . OR G 16 | A NA LY T I CS - MAGA Z I NE . OR G
branch offces in response to the rapidly
changing economy in China. Still another
describes how the Federal Aviation Ad-
ministration used analytics to manage
air traffc more safely and cost effectively
during periods of bad weather.
It’s all useful information for learning
about what can be done with analytics,
but it’s only a precursor to the central fo-
cus of the site, which is how to start us-
ing analytics. Guidelines are presented
on how to recognize an opportunity that
could beneft from analytics. The site
also provides information on structuring
an analytics project, working with an an-
alytics professional, and fnding the right
analytics professional for the job at hand.
Analytics is, after all, a broad discipline.
The site provides a list of companies
and consultants that can help with ana-
lytics initiatives. Unfortunately, the list is
a work in progress, built on a database
in need of updating. The site also lacks
functionality that would allow companies
and consultants to update their informa-
tion directly. INFORMS is working on it.
As an alternative, individuals are invited
to submit an online form describing what
analytics activities they’re considering. The
forms are reviewed by a panel, which then
assigns an “analytics mentor.” The mentor
then contacts the individual and provides as-
sistance getting started. The service is free
of charge, though the goal is to point the in-
dividual in the right direction, not to perform
an extended consulting project. Individuals
may go to www.informs.org/analyticsmentor
and submit the form to request help. An IN-
FORMS staffer will be in touch quickly.
I’m not sure which is harder when it
comes to analytics: deciding to take a
frst step or actually taking it, but in either
case INFORMS’ new website has a lot to
offer. It’s certainly not the same as work-
ing side by side with capable, supportive
people, but it can make a foreign land
look a little less foreign. And the mentors
program provides a direct link into a com-
munity of knowledgeable analytics prac-
titioners. Analytics brings tremendous
value, but it’s also new and exciting. Why
wait to take the frst step? ❙
Andrew Boyd, senior INFORMS member and
INFORMS VP of Marketing, Communications and
Outreach, was involved in the development of
the website described in this column. He can be
reached at [email protected]
PROFI T CENTER
Request a no-obligation INFORMS Member Benefits Packet
For more information, visit: http://www.informs.org/Membership
sas.com/vademo
Experience Visual Analytics frsthand.
Explore billions of rows of data in minutes or seconds, visually represented in a way that brings hidden patterns
into plain sight. Then easily create and share reports on the Web or mobile devices.
See your data for all it’s worth.
VISUAL
SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. ® indicates USA registration. Other brand and product names are trademarks of their respective companies. © 2013 SAS Institute Inc.
All rights reserved. S105651US.0213
WWW. I NF OR MS . OR G 18 | A NA LY T I CS - MAGA Z I NE . OR G
As described in the previous edition of Analyze
This!, I am currently working on a research study with
Jeanne Harris at Accenture’s Institute for High Per-
formance. Specifcally, we are seeking to develop a
quantitative and qualitative understanding of the cur-
rent state of analytics practice. If you have already
completed the on-line survey, please accept our grati-
tude (extra kudos to those who have sent us emails
with additional thoughts and ideas).
For the rest of you, it is most defnitely NOT too
late: just click here to provide your valuable input into
our research process. Full disclosure: Based on the
frst wave of respondents, the mean time to complete
the survey is just under 15 minutes.
In our study, we are trying to practice what we
preach by gathering and analyzing data to empirically
test some of our hypotheses about what we think is
going on. Meanwhile, the “big data” and “analytics”
hype is almost deafening. In February 2013, Gartner
changed the name of their celebrated Magic Quad-
rant from “Business Intelligence” to “Business Intel-
ligence and Analytics Platforms” [1], while formally
including predictive modeling, data mining and inter-
active visualization into its evaluation criteria. Over at
LinkedIn.com, there are more than 2,500 analytics-
related groups, including eight with more than 20,000
members and 30 with more than 5,000 members.
Are analytics and big data
overhyped?
BY VIJAY MEHROTRA
One reason that the
hype … sounds so
familiar may be that
many of the same
players from the last
revolution are also a
part of this one.
ANALYZE THI S!
MAR CH / A P R I L 2013 | 19 A NA L Y T I C S
While less than fve graduate programs
in analytics existed in 2011 in the United
States, there will be at least 17 such pro-
grams as of fall 2013 [2]. Meanwhile, an-
alytics-oriented headlines seem to crop
up every day in publications as diverse
as InfoWorld [3], the New York Times [4],
Atlantic Monthly [5] and Forbes [6].
WHAT REALLY MATTERS
All of this was on my mind recently
when I happened across a May 2003
Harvard Business Review article entitled
“IT Doesn’t Matter,” written by Nicholas
Carr (who later expanded it into a book
[7]). Carr’s basic thesis was that despite
all of the hype, information technology
actually no longer provided its adopters
with competitive advantage. Placing IT
along a continuum of innovations such
as steam shovels, trains and electricity,
Carr argued that the economics of such
technological breakthroughs – and in
particular, the need for those that devel-
oped them to sell them widely in order to
recoup their development costs – meant
lower prices and rapid commodifcation.
Given the visibility of the article and
its author (Carr was an editor-at-large
at HBR at the time) and its provocative
title, the article was met with a frestorm
of criticism. Several letters to the editor
were subsequently published. Steven Al-
ter, who is currently a colleague of mine,
thoughtfully pointed out that information
systems are merely a component of what
he calls “work systems” [8]; because IT
is an essential component of most mod-
ern corporate work systems, it did indeed
matter a great deal.
Others responses were less san-
guine. Gartner executives Marianne
Broadbent, Mark McDonald and Richard
Hunter warned that, “the danger is that
by scanting the fantastic potential for in-
novation that lies ahead in IT, Carr will
lead executives to focus only on control-
ling IT costs.” Paul Strassman, former
CIO of General Foods and Kraft, wrote
of Carr that, “he bases his conclusions
entirely on his reasoning, by analogy…
any proof that rests entirely on analogies
is fawed. This technique was used to up-
hold medieval dogma, and it delayed the
advancement of science by centuries.”
NEW ALBUM, SAME BAND
After reading this article and the as-
sociated letters, a number of thoughts
struck me:
• One reason that the hype
surrounding big data and analytics
sounds so familiar may be that many
of the same players from the last
revolution are also a part of this one.
For example, see Accenture, Gartner
and InfoWorld above, as well as
companies like IBM and Oracle that
WWW. I NF OR MS . OR G 20 | A NA LY T I CS - MAGA Z I NE . OR G
have been aggressively acquiring
analytics companies over the last
few years. Yeah, the band’s defnitely
got a new album out, but there’s no
mistaking the sound.
• Software vendors, new ones as well
as old, are working furiously to bring
analytics software products to market.
But these vendors should beware of
making extravagant promises that they
are likely to fail to deliver on. As John
Seeley Brown (former chief scientist
at Xerox PARC) and John Hagel III
(former McKinsey principal) pointed out
in their response to Carr’s article:
“Rather than help companies
understand that IT is only a tool,
technology vendors have tended to
present it as a panacea. ‘Buy this
technology and all your problems will
be solved.’…When the anticipated
results did not materialize, the
backlash began to gather in
executive suites … ‘let’s buy as little
as we can and squeeze the vendors
as much as we can.’ ”
• Along these same lines, the new
offered wisdom is that the real value
of previous IT expenditures (many of
which have heretofore been viewed
as bad investments) is the data they
are providing to support analytics.
Part of the sales pitch for analytics
and big data today is that managers
and executives need to jump on the
big data train in order to capture
those long-ago promised benefts. In
Vegas terms, they are being asked to
“double down.”
• In the end, we are probably best off
to think of IT, big data and analytics
as part of a classical supply chain
whose mission is to capture, store
and analyze data and then interpret,
communicate and utilize the results
to deliver business insights, make
better decisions, and ultimately
achieve increased proftability with
decreased risk. Regardless of where
the media might be focused at
any given time, the results largely
depend on the capability of the
weakest link.
Indeed, another key (and rarely men-
tioned) factor in the increased accep-
tance and utilization of analytics in the
business world is the ascendance of a
new generation of managers and execu-
tives – a generation that has grown up
with computers and is more comfortable
with what they can and cannot do.
Whoops! That last paragraph sounds
like more big data hype! Please note: This
claim about more enlightened managers
and executives is just a hypothesis, one of
several that we seek to examine in the re-
search project that I talked about at the top
ANALYZE THI S!
MAR CH / A P R I L 2013 | 21 A NA L Y T I C S
of this column. To test these hypotheses,
we need data! Please fll out our survey and
help us to better understand what’s really
going on out there in the world of analytics.
We promise to share our results with all of
you, both here and elsewhere, to help you
cut through all the noise, which seems to be
getting louder every day. ❙
Vijay Mehrotra ([email protected]) is
an associate professor in the Department of
Analytics and Technology at the University of San
Francisco’s School of Management. He is also an
experienced analytics consultant and entrepreneur,
an angel investor in several successful analytics
companies and a longtime member of INFORMS.
http://jps.informs.org
Attention Companies Who Hire Analytics Professionals
Reserve your space now
for the industry’s premier,
professional job fair!
Brought to you by Analytics Magazine
 Find the seasoned professionals you need – over 700 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 ORMS Today
[email protected] or call (800) 4-INFORMs
Questions?
Analytics Connect
April 7-9, 2013 | Grand Hyatt San Antonio | San Antonio, Texas
Applying Science to the Art of Business
USINESS ANALYTICS &
PERATIONS RESEARCH
INFORMS CONFERENCE ON
REFERENCES
1. Click here for more details on the Magic Quadrant.
2. For a listing and additional details about
graduate programs in analytics, click here.
3. “Why You Should Jump Into Big Data” is
accessible online. To view it, click here.
4. “The Origins of ‘Big Data’: An Etymological
Detective Story” is accessible online. To view it,
click here.
5 “ Can Big Data Save American Schools?” is
accessible online. To view it, click here.
6. “Revolution in the Big Data and Business
Landscape” is accessible online. To view it, click here.
7. “Does It Matter” is available through Amazon.
For details, click here.
8. For more, see Alter’s book, “The Work System
Method. For details, click here.
WWW. I NF OR MS . OR G 22 | A NA LY T I CS - MAGA Z I NE . OR G
The Institute for Operations Re-
search and the Management Sciences
(INFORMS), the premier source for ad-
vanced analytics (and the publishers of
Analytics magazine), will administer the
inaugural exam for its Certifed Analytics
Professional (CAP™) program April 7 in
conjunction with the INFORMS Confer-
ence on Business Analytics & O.R. in San
Antonio, Texas.
The CAP™ program, designed to be
the industry’s gold seal of approval for an-
alytics professionals, consists of several
elements, including a 100-question, mul-
tiple-choice exam. In order to be eligible
to take the exam, candidates must have
a bachelor’s degree or higher, between
three and seven years of analytics work-
related experience (depending on de-
gree) and verifcation of communication
and other “soft skills,” such as effective
Certified Analytics
Professional
CAREER BUI LDER
partnering with business clients, framing
problems with stakeholders, working in
project teams and communicating results
to decision-makers.
Exam questions will cover such domains
as problem framing, data management,
methodology selection, model building, de-
ployment and lifecycle management.
Three other exam sites are also
scheduled for 2013: June 23 in Chicago
(preceding the INFORMS Healthcare
Conference), July 13 in McLean, Va.
(Booz Allen Hamilton offces) and Oct. 5
in Minneapolis (preceding the INFORMS
Annual Meeting).
For further information regarding the
CAP™ program, including a download-
able “Candidate Handbook” (with sample
exam questions) and certifcation appli-
cation and agreement, click here. ❙
Subscribe to Analytics
It’s fast, it’s easy and it’s FREE!
Just visit: http://analytics.informs.org/
APRIL 7, JUNE 23, JULY 13, & OCTOBER 5, 2013
Conducted BY INFORMS, the leading professional society in advanced analytics
INAUGURAL ANALYTICS CERTIFICATION EXAMS

Be among the first to become a Certified Analytics Professional (CAP

).
Make plans now to take the profession's first analytics certification exam.
Candidate Handbook now available at
www.informs.org/Build-Your-Career/Analytics-Certification.
BENEFITS OF CERTIFICATION
• Advances your career potential by setting you apart from the competition
• Drives personal satisfaction of accomplishing a key career milestone
• Helps improve your overall job performance by stressing continuing
professional development
• Recognizes that you have invested in your analytics career by pursuing this rigorous credential
• Boosts your salary potential by being viewed as experienced analytics professional
• Shows competence in the principles and practices of analytics
ELIGIBILITY
• BA/BS or MA/MS degree or higher
• At least five years of analytics work-related experience for BA/BS holder in a related area
• At least three years of analytics work-related experience for MA/MS (or higher) holder in
a related area
• At least seven years of analytics work-related experience
for BA/BS (or higher) holder in an unrelated area
APPLICATIONS
• Open in January, 2013
• Prepare to apply by reviewing Candidate Handbook now
• Arrange now to secure academic transcript and
confirmation of “soft skills” from employer
to send to INFORMS
COST
• $495 INFORMS Members
• $695 Non-Members
• Bundled rates with meetings available
QUESTIONS
• Email [email protected]
2013 CAP

EXAM SCHEDULE
APRIL 7, 2013
INFORMS Conference on
Business Analytics & O.R.,
San Antonio, TX
DOMAINS OF ANALYTICS PRACTICE
Domain Description Weight*
Business Problem (Question) Framing
Analytics Problem Framing
Data
Methodology (Approach) Selection
Model Building
Deployment
Life Cycle Management
*Percentage of questions in exam
I
II
III
IV
V
VI
VII
15%
17%
22%
15%
16%
9%
6%
100%
OCTOBER 5, 2013
INFORMS Annual
Meeting,
Minneapolis, MN
JUNE 23, 2013
INFORMS Healthcare
Conference,
Chicago, IL
July 13, 2013
Booz Allen Hamilton
Offices,
McLean, VA
TM
WWW. I NF OR MS . OR G 24 | A NA LY T I CS - MAGA Z I NE . OR G
The promise of technology-enabled population
health management.
Smart Care
BY (LEFT TO RIGHT)
RAJIB GHOSH, KYLIE M. GRENIER
AND THEO AHADOME
HEALTHCARE ANALYTI CS
MAR CH / A P R I L 2013 | 25 A NA L Y T I C S
For many months after its
passage by the U.S. Con-
gress, the Affordable Care
Act (ACA) witnessed a grow-
ing sense of uncertainty over its future.
Stakeholders in the U.S. healthcare val-
ue chain didn’t take much action. Finally,
two successive events in 2012 brought
back the momentum. First, the U.S. Su-
preme Court upheld the constitutional-
ity of the act. Then the U.S. presidential
election results made sure that the act
would stay and accelerate. Business
stakeholders and millions of patients that
can beneft from the ACA are now looking
for the path forward.
Pay for performance and value-based
purchasing are the key pillars of the ACA,
which aspires for the “triple aim”: better
quality of care for population, better ex-
perience of care for individuals and lower
per capita cost for care delivery. This ar-
ticle focuses on opportunities, barriers
and technologies available to enable
population health management (PHM),
which we believe will build the foundation
for the rest of ACA’s objectives.
Population health management (PHM)
is a holistic approach to healthcare that aims
to improve the health of an entire popula-
tion. Along with the focus on medical care
for the population, PHM also looks at reduc-
ing existing health inequities and the social
determinants, such as environment, social
structure, resource distribution, etc. at the
population level.
PHM is instrumental for the suc-
cessful implementation of the ACA. ACA
opens the door to an estimated 32 mil-
lion more uninsured U.S. citizens to join
an already stretched healthcare system
by 2014. Additionally, at the current rates
of physicians graduating and retiring, the
United States could experience a short-
age of 150,000 physicians in the next 15
years. Given this chasm of supply and de-
mand, the U.S. healthcare system needs
to focus on three key components as de-
scribed by the Care Continuum Alliance:
the central care delivery and leadership
roles of the primary care physician; the
critical importance of patient empower-
ment to accept personal responsibility in
their care; and the expansion of care co-
ordination through wellness, disease and
chronic care management. Those are
also the core building blocks for PHM. To
achieve that, disparate IT systems need
to interoperate, and powerful predictive
analytical models need to be employed
at a population level. Such large promis-
es naturally come with large challenges.
CHALLENGE OF POPULATION
HEALTH MANAGEMENT –
A CASE STUDY
For a PHM to work, multiple disparate
healthcare systems governed by both public
F
WWW. I NF OR MS . OR G 26 | A NA LY T I CS - MAGA Z I NE . OR G
and private entities have to contribute data
to one master client record. Additionally,
that data must also be accessible to the cli-
ent. Let’s take a case study to illustrate the
intricacies involved in the process. Figure 1
shows some of the potential data sources
for San Francisco, with a popula-
tion of 800,000 people.
There are many subcircles
within each circle, which in
turn could contain hundreds
of standalone databases with
confdential medical data, not
to mention myriad paper fles
yet to be digitized. This diagram
represents only the segment
of the population that seeks
medical care – an estimated 27
percent [1]. That means only 27 percent
of the symptomatic population is known
to one of the healthcare entities in San
Francisco.
To begin the challenge of bringing PHM
data together for the 27 percent seeking
healthcare, a system inventory prepared by
the project team at the University of Califor-
nia San Francisco (UCSF) Medical Center
catalogued the number of inbound and out-
bound interface requests within their enter-
prise (Figure 2).
In this example, 37 percent of the
available data was out of scope for the
initial phase of the project for a number
of technical reasons. In addition, UCSF
is only one of 21 possible entities that
have data on the client that could go into
the medical record. So, building a com-
prehensive database for each individual
client will take new and creative thinking
to overcome several challenges:
POPULATI ON HEALTH MANAGEMENT
Figure 1: PHM radial cycle – health systems in
San Francisco.
Figure 2: Inter-system communications at the University
of California at San Francisco (UCSF) Medical Center.
MAR CH / A P R I L 2013 | 27
A NA L Y T I C S
1. Partnerships with diverse public and
private agencies
2. Data security and privacy combined
with client access
3. Proprietary database systems and
vendor partnerships
4. Available resources for a project of
this magnitude: both budget and
skilled technical resources
5. Evolving healthcare employee roles,
shortage of nurses and other primary
care resources
6. Rapidly changing technology tools
A UCSF healthcare integration man-
ager described the healthcare data co-
nundrum as perhaps beyond the reach
of a relational database. Standards for
commerce data are quite different than
the standards for healthcare data, and
that adds another layer of complexity. To
scale this task and begin to understand
the unknown factors unique to health-
care, perhaps it’s time to consider thinking
beyond both traditional database design
models and current healthcare delivery
models. Those are daunting challenges,
April 7-9, 2013
Grand Hyatt San Antonio
San Antonio, TX
Join executives from Boeing, IBM, Intel, Chevron, Nielsen, IRS,
Ford, Schneider, Target, Land O’Lakes, UPS, Bank of America, HP,
Verizon, Jeppesen, Verizon Wireless, and more!
This special program for senior
executives provides decision
makers with a bottom-line
understanding of how
business analytics & O.R.
are used to drive better
organizational decisions.
• special networking reception
hosted by the INFORMS
Board of Directors,
• executive-level panel discussion,
• selected executive-level talks.
Submit your registration at
meetings.informs.org/analytics2013/execforum.html
4
th
Annual Executive Forum
Nominate Your Executives to Attend
Applying Science to the Art of Business
USINESS ANALYTICS &
PERATIONS RESEARCH
INFORMS CONFERENCE ON
WWW. I NF OR MS . OR G 28 | A NA LY T I CS - MAGA Z I NE . OR G
but some IT solutions are currently avail-
able to help.
TECHNOLOGY LANDSCAPE
IN POPULATION HEALTH
MANAGEMENT
As evident from the San Francisco
case study, a complex number of sys-
tems from different facilities need to
be connected to achieve a functional
population health management system.
These can essentially be broken down
into two subsets: 1. connecting systems
within one facility, and 2. interconnecting
systems between facilities.
Within a single facility, the electronic
health record (EHR) serves as a key ag-
gregator of patient records from across
departments. The EHR becomes a hub
for inbound and outbound data functions.
Some of the systems it interacts with
include fnancial and administrative
systems, labs, radiology, cardiology, neo-
natal, ICU and obstetrics. As illustrated in
the case study, this can result in myriad
interconnections with data in proprietary
formats within departmental silos. Some
technology solutions, however, are avail-
able to alleviate this problem. For image-
based data, a vendor-neutral archive
(VNA) can mitigate enterprise commu-
nication by serving as a central hub for
images. For all other communications to
the EMR, an enterprise content manager
(ECM) can similarly aggregate data to
manage solicited and unsolicited com-
munications from disparate systems, en-
suring that individual systems need not
directly communicate with the EHR or
vice-versa.
Outside of the single fa-
cility, systems from home
health, wellness centers and
primary care, among others,
need interconnection within
the context of an accountable
care organization (ACO),
integrated delivery network
(IDN) or other functional
inter-facility relationship. As
in UCSF’s PHM, those can
communicate directly with
a facility’s EHR (McKesson
Home Health to Epic EHR as
POPULATI ON HEALTH MANAGEMENT
Figure 3: Population health management begins at the single
facility level.
* Others include nurse station, pharmacy, maternity,
discharge and back offce
MAR CH / A P R I L 2013 | 29
A NA L Y T I C S
illustrated) or with the enter-
prise content manager at the
receiving facility. As ACOs,
IDNs and non-related facili-
ties need interconnection, a
health information exchange
(HIE) is formed to allow such
communications. PHM be-
comes the management of
patient fow within those in-
terconnected systems.
The rapidly grow-
ing level of technology
adoption in different department and
sub-systems of the healthcare entities
compounds the challenge of central-
ized data management. According to In-
Medica, by 2016, more than 80 percent
of healthcare facilities will have picture
archiving and communication systems
(PACS), radiology information systems
(RIS), critical care information systems,
obstetrical information systems and
EMRs, with the majority of these systems
approaching full saturation.
The challenge of rising IT adoption
also presents an opportunity for achiev-
ing PHM – departments and facilities that
may not have been able to share infor-
mation previously are able to do so as
they each become more digitized.
As these technologies are increas-
ingly adopted, so, too, does the need
to feasibly manage and share the rising
volume of patient data they store. In the
imaging space, vendor-neutral archives
are already taking off with more than 10
percent saturation level in the United
States in 2012. Other integration plat-
forms such as enterprise content manag-
ers and HIEs are still far below this level.
For hospital IT, it is important to identify
these growing data sources and imple-
ment appropriate data management sys-
tems for imaging and medical records.
THE ROLE OF PAYERS IN
POPULATION HEALTH MANAGEMENT
Both InMedica and IDC are predicting
that this year providers will look at utiliz-
ing healthcare IT technologies to accel-
erate their compliance with “Meaningful
Use Stage 2,” revenue cycle manage-
ment and optimizing analytic resources.
Those are critical to handling the bottom
Figure 4: Toward PHM – adoption in the United States (level of
market penetration).
Source: InMedica
WWW. I NF OR MS . OR G 30 | A NA LY T I CS - MAGA Z I NE . OR G
line impact, which may be impacted by hospital read-
mission penalties or value-based purchasing models.
Therefore, enabling PHM, albeit important for triple
aim, may take a back seat. However, currently 64
percent of the ACO programs are governed through
a joint partnership between payers and providers,
which is expected to grow in 2013. Payers need to
play a critical role to enable their provider partners
to prepare and roll out PHM fundamentals. They will
do so through the use of data liquidity, digital health
solutions and analytics-as-a-service. In a recent in-
terview published in HealthLeaders Media, Reed
Tuckson, M.D., executive vice president and chief
of medical affairs of United Healthcare, expressed a
similar view.
Despite the fact that the concept and benefits
of PHM are neither new nor revolutionary, payers
for many decades shunned the role of being the
leader in deploying this approach at scale. There
are, however, examples of provider-led and limit-
ed-scope PHM in action in some parts of United
States. Monarch Healthcare, a physician led and
owned independent practice association (IPA) in
Orange County (California) [3] has demonstrated
success in their PHM efforts. However, a broader
and at-scale deployment requires more muscle
power in technology, data acquisition and finan-
cial resources. This is where payers are expected
to play a pivotal role. But why would they do this
POPULATI ON HEALTH MANAGEMENT
Join the Analytics Section of INFORMS
For more information, visit:
http://www.informs.org/Community/Analytics/Membership
Payers need to play a
critical role to enable
their provider partners to
prepare and roll out PHM
fundamentals. They will do
so through the use of data
liquidity, digital health
solutions and analytics-as-
a-service.
www.tradeextensions.com [email protected]
Corporate Headquarters U.S. Headquarters Europe Headquarters
Uppsala, Sweden +46 (0) 18 13 66 00 Houston, TX USA +1 (855) 215 8387 Swafham, UK +44 (0)1760 720 746
Trade Extensions has the technology and expertise to enable businesses and
sourcing professionals to make better sourcing decisions – decisions that are
strategically-aligned and optimized against operational constraints.
The Trade Extensions sourcing platform is used by some of the world’s most respected
organizations including: P&G, Coca-Cola, Unilever, BP,
Kimberly-Clark, Cargill, Impress, Alcan, Dow, Cabot, Huhtamaki,
Ineos Polyolefins and AT Kearney.
WWW. I NF OR MS . OR G 32 | A NA LY T I CS - MAGA Z I NE . OR G
POPULATI ON HEALTH MANAGEMENT
system provider Active Health and iTri-
age (a provider finder app for emergen-
cy situations) and United’s acquisition
of Optum and Ingenix are testaments
to this shift in paradigm.
FROM PHM TO SMART CARE
PHM allows a complete view of a
patient’s health from all avenues of
care and ensures that caregivers are
able to utilize this to understand a pa-
tient’s full care pathway. “Smart Care”
is the embodiment of “smart” systems
to improve patient flow through intel-
ligent interchange of data. Smart Care
means patient data is immediately
forwarded from one department to
another or from one facility to anoth-
er depending on where the patient is
scheduled to go next. It means data is
immediately available at the right place
and in the right format, eliminating de-
lays in identifying patient records. It
means patients can be directed to fa-
cilities with lower utilization rates when
a potential overburden on another fa-
cility is detected based on likely flow
of patients from test results. It saves
costs and enhances productivity.
Smart Care also includes analyt-
ics – the ability to analyze and identify
trends in patient data and recommend
intervention before exacerbation of
a condition. A patient visits a fitness
now? The short answer is to improve
their medical loss ratio (MLR) and to
acquire trust and engagement of the
estimated 32 million new customers
– many will not be acquired through
institutional buyers of the past such
as employers or brokers but as direct
consumers via the health insurance
marketplaces or exchanges.
This is a paradigm shift in the pay-
er industry that has never happened
before. Payers will not like to see sev-
eral of their new customers show up
at the emergency room when their ex-
acerbations can actually be prevented
through PHM. If payers can enable
their provider partners who are jointly
governing ACOs to successfully de-
ploy PHM, they can impact top-line
revenue through retention and in-
creased member acquisition and lower
their risk exposure. That is a winning
strategy. During the last two years, the
healthcare industry has seen signifi-
cant investment from the leading pay-
ers to build the technology portfolio
to achieve this. Aetna’s purchase of
HIE vendor Medicity, decision-support
Subscribe to Analytics
It’s fast, it’s easy and it’s FREE!
Just visit: http://analytics.informs.org/
MAR CH / A P R I L 2013 | 33
A NA L Y T I C S
coach who measures the patient’s
weight. Six months later, upon a revisit
to the same fitness coach, the weight
loss is so dramatic that the fitness
coach alerts a primary caregiver. It
may or may not be too late. However,
imagine a scenario where the patient
does not visit the fitness coach again
within the six-month period. Instead
he enrolls in a corporate wellness pro-
gram in month three. In Smart Care,
the weight measured by the fitness
coach is available, and rapid weight
loss is determined and acted upon,
even though data is from different fa-
cilities. The caregiver does not need to
notice this trend; the system does. The
interconnection between the two facili-
ties is achieved by the PHM – Smart
Care takes this further and acts upon
the shared data.
Population health management is the
cornerstone of Smart Care – the 21st
century technology enabled coordinated
care across the nation. With the legisla-
tive mandates of ACA providing the im-
petus, PHM has the best opportunity to
succeed. The path forward is not devoid
of barriers, but the good news is there
is a path in sight and stakeholders are
incentivized to do this to serve their own
good, thus creating a virtuous cycle.
Technology and standards are catching
up, albeit the latter is somewhat slower
than the former. Many oars are in the wa-
ter, and for the frst time it seems they are
all rowing in the same direction. ❙
Rajib Ghosh ([email protected]) has 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 held senior
positions at Hill-Rom, Solta Medical and Bosch
Healthcare. He is an independent consultant and
business advisor. His recent work interest includes
public health and the feld of health IT enabled
sustainable healthcare delivery.
Kylie M. Grenier ([email protected]) is
a career technologist with 30 years of experience
transforming business. She is currently leading
change at the U.C. San Francisco School of
Nursing – a top three medical graduate school and
nationally ranked research program.
Theo Ahadome ([email protected]) is
a senior market analyst at InMedica, the medical
research group of IMS Research (recently acquired
by IHS). Ahadome is the company’s primary analyst
in the telehealth and healthcare IT research areas.
He has published a number of industry-acclaimed
reports on the telehealth, consumer medical and
healthcare IT markets. Ahadome regularly presents
on trends in these markets and in the healthcare
sector at supplier and industry events worldwide.
REFERENCES
1. L.A. Green, et al., 2001, “The Ecology of
Medical Care Revisited,” New England Journal of
Medicine, Vol., 344, No. 26, June 2001.
2. Kerr L. White, et al., 1961, “The Ecology of
Medical Care,” 1961, Journal of Urban Health,
Vol. 265, No. 18.
3. Kathleen L. Carluzzo, et al., 2012, “Monarch
HealthCare: Leveraging Expertise in Population
Health Management,” The Commonwealth Fund,
January 2012.
WWW. I NF OR MS . OR G 34 | A NA LY T I CS - MAGA Z I NE . OR G
Barriers to their deployment are coming down while
opportunities for their deployment are improving.
Recommendation
engines at work
BY CHRISTOPHER BERRY
ONLI NE MARKETI NG
MAR CH / A P R I L 2013 | 35 A NA L Y T I C S
If you’ve used the stream-
ing service of Netfix, bought
something from Amazon or
connected with “people you
may know” on LinkedIn or Facebook,
then you’ve used a recommendation en-
gine. And chances are, you’ve watched
more movies, bought more stuff and are
less likely to churn because of them.
These engines work.
Recommendation engines match
features about you with things that you
might be interested in.
For instance, a movie has a release
year, a genre, actors and box offce re-
sults. You have features. You have pref-
erences, an age, and you may have
completed a survey expressing some of
your attitudes toward certain movies. You
may have rated some of the movies you
watched. By fguring out which sets of
movies to show you, and your response
to those recommendations, the machine
learns over time to make better sugges-
tions. If you watched a few science fction
movies, and you rated them highly, then
the engine will learn to show you more sci-
ence fction movies, and, for variety, mov-
ies that other people like you, who like
science fction movies, might also enjoy.
More and more companies are using
recommendation engines. Apple has its
own engine to help consumers fnd apps
they are likely to enjoy from Apple’s large
inventory. Microsoft’s XBOX 360 Live
has an engine to suggest new games
you might be interested in based on what
you’ve previously shown an interest in.
Many of the algorithms that are used
in recommendation engines and ma-
chine learning aren’t all that new. Regres-
sion, decision trees, K-nearest neighbor
(KNN), support vector machines (SVMs),
neural networks and naive Bayes are
established methods with well-known
constraints and appropriate uses. Many
of these methods have been used to
support data-driven business decision-
making for a long time. So, if the benefts
of recommendation engines have been
long known, what’s different now? What’s
causing more companies to implement
recommendation engines to support cus-
tomer decision-making?
THREE COST TRENDS
Three major trends are driving the shift
from possible to scalable and enabling
recommendation engines to scale tech-
nologically, economically and effectively.
1. The cost of data storage has come
down. $500 buys a large volume of
space on Amazon’s cloud. It’s the equiv-
alent of what would have cost hundreds
of thousands of dollars just 10 years
ago. In short, big data also means cheap
storage.
I
WWW. I NF OR MS . OR G 36 | A NA LY T I CS - MAGA Z I NE . OR G
2. The cost of software has also come
down dramatically. The software that
enables companies to manage a large
amount of data used to, and still does,
run into the millions of dollars. Howev-
er, thanks to decisions by several com-
panies to open source their software,
programs like Hadoop and Druid are
monetarily free. While it certainly takes
experts time to set it up and maintain, the
overall cost of ownership has fallen. This
abstraction has enabled smaller teams
to tackle much bigger opportunities, like
recommendation engines.
3. The cost of data has also come
down. People themselves, in part driven
by smartphone adoption, emit large vol-
umes of storable data. Some of this data
is very unstructured and dirty, compa-
rable to call center data. Some is clean,
like GPS location data. Moreover, it has
become popular for start-ups to offer ap-
plication programming interfaces (APIs).
So not only is more data generated in
more places, but it’s more available to
more people in more places.
Falling barriers herald bold experi-
mentation. These three trends intersect
and cause a Cambrian explosion of ex-
perimentation and commercialization
attempts.
INTRODUCING THE DATA SCIENTIST
At the center of this is the data scien-
tist. Data scientists turn data into product.
A recommendation engine is certainly a
product.
Data scientists combine a number of
skills. They have to know how to write
code. They know statistics and the algo-
rithms used to extract patterns from na-
ture. And they understand business. The
combination of the three skills increas-
es the likelihood that their solutions will
scale successfully.
Data scientists come from any num-
ber of backgrounds. Some are highly ac-
complished computer scientists that got
deeper into business and statistics. Some
are from biomedical informatics. Many are
from sectors with particularly high numer-
acy, such as physics. Others have roots
in the management sciences. There are
a number of ways a person can level up
to become a data scientist, but it gener-
ally ends with a data scientist possessing
competence in all three skill sets.
Data science begins with data. Noth-
ing gets built without data. Data sci-
ence continues with science. Accurate,
RECOMMENDATI ON ENGI NES
Help Promote Analytics
It’s fast and it’s easy! Visit:
http://analytics.informs.org/button.html
MASTER OF SCIENCE IN ANALYTICS
• 15-month, full-time, on-campus program
• Integrates data science, information technology and business applications
into three areas of data analysis: predictive (forecasting), descriptive (business
intelligence and data mining) and prescriptive (optimization and simulation)
• Offered by the McCormick School of Engineering and Applied Science
www.analytics.northwestern.edu
MASTER OF SCIENCE IN PREDICTIVE ANALYTICS
• Online, part-time program
• Builds expertise in advanced analytics, data mining, database management,
fnancial analysis, predictive modeling, quantitative reasoning, and web analytics,
as well as advanced communication and leadership
• Offered by Northwestern University School of Continuing Studies
877-664-3347 | www.predictive-analytics.northwestern.edu/info
NORTHWESTERN ANALYTICS
As businesses seek to maximize the value of vast new streams of available data,
Northwestern University offers two master’s degree programs in analytics that
prepare students to meet the growing demand for data-driven leadership and
problem solving. Graduates develop a robust technical foundation to guide
data-driven decision making and innovation, as well as the strategic,
communication and management skills that position them for leadership roles
in a wide range of industries and disciplines.
WWW. I NF OR MS . OR G 38 | A NA LY T I CS - MAGA Z I NE . OR G
persuasive and effective prediction requires patterns.
The process of discovering that pattern is science.
Any product worth building requires a reliable pattern
to exist in the data.
The process of exploiting that pattern, especially
for commercial gain, is engineering. Data science
generally ends with engineering.
Many people who work with data scientists, who
are responsible for various aspects of product build-
ing, are engineers. They may have data science in
their titles. While they are likely to confuse human
resource departments and leadership alike, the am-
biguity is well worth the cost. The engineers are indis-
pensable for translating the patterns found in nature
directly into business outcomes. This ambiguity is a
cause for concern among those concerned solely
with labels.
The output of data science is product. As a result,
product management is a major concern. Because
their stance is based frmly in science and in itera-
tion, data scientists frequently chose methods and
tools that emphasize iteration and experimentation.
Ideas such as fast-failure and continuous deploy-
ment are particularly well suited to this type of prod-
uct development. When data scientists maintain their
own product management and development teams,
they chose continuous deployment, agile methods
and rapid iteration.
RECOMMENDATI ON ENGI NES
Any product worth
building requires a
reliable pattern to
exist in the data. The
process of exploiting
that pattern, especially
for commercial gain,
is engineering. Data
science generally ends
with engineering.
Request a no-obligation INFORMS Member Benefits Packet
For more information, visit: http://www.informs.org/Membership
MAR CH / A P R I L 2013 | 39
A NA L Y T I C S
The difference in stances is likely to
cause cultural tension within organiza-
tions. The tension may generate positive
spin-offs so long as it does not generate
regrettable churn.
HOW DATA SCIENTISTS WORK
Data scientists spend time under-
standing the metric that needs to be
maximized and the business context for
that maximization. They call this the op-
timization objective and remain focused
on a single one.
Like their cousins in operations, data
scientists frequently have to gather data
into one place. If there is none available,
or if the odds of unlocking existing data
are too remote, they have to generate
their own source. Those from the natural
sciences will gravitate toward setting up
an experiment to get some data on which
to train an engine.
Data scientists avoid writing as
much of their own code as possible.
They use open source libraries from Py-
thon, Octave and R before they resort
to over-optimization. They will sooner
use Amazon Mechanical Turk to obtain
a larger data set than invent their own
framework.
They will think about which meth-
ods are likely to scale, and they will
try them out. They separate their data
into a training set (in-sample set), a
cross-validation set and a testing set
(out-of-sample set). They will try to
avoid over-fitting or under-fitting their
algorithm to the data. They will seek
a compromise between recall and
precision.
They’ll expose their recommendation
engine to the wild, observe how people
react to it and then use that information to
refne its accuracy. They will keep the en-
gine out there as they gradually improve
it. They’re rarely done optimizing both
its scale and its performance, frequently
seeking out additional data streams to
use to improve it.
IMPACT
Recommendation engines put data
immediately to work for the business and
for consumers. The barriers to their use
have come down and the opportunities
for their deployment have improved. As
more and more companies are discov-
ering, they cost a small fraction of an
average advertising campaign, bring in
directly attributable revenue and deliv-
er surprisingly short payback periods if
done right. ❙
Christopher Berry ([email protected]) is the
co-founder and chief science officer of Authintic
(www.authintic.com), an analytics technology
company based in Toronto, Canada. Prior to
Authintic, Berry built the measurement science
and labs groups at Syncapse, a social media
technology company, and the marketing science
department at Critical Mass.
April 7-9, 2013
Grand Hyatt
San Antonio,
Texas
Showcasing
THE POWER
OF
ANALYTICS
BOEING
IBM
TARGET
GOOGLE
MIT
INTEL
DELL
USAA
GARTNER
STARBUCKS
FORD
NEILSEN
STANFORD
BOOZ ALLEN
HAMILTON
CHEVRON
SAS
PEPSICO
TERADATA
GE
KROGER
DELOITTE
and 70 other top companies,
universities and government agencies
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.
• Real-world use of descriptive, PREDICTIVE and prescriptive analytics
• Focus on BIG DATA, marketing analytics, forecasting
• Most rigorous and REAL-WORLD analytics conference offered
• Administration of Certified Analytics Professional exam
• Not just what to do but HOW TO DO IT
Keynote Speakers
Edelman Award Ceremony & Banquet
Nicole Piasecki
Vice President,
Business Development
& Strategic Integration
Boeing Commercial Air-
planes
Sandy Carter
Vice President
IBM Social Business
Evangelism
and Sales
IBM
Learn from the BEST about High-Impact
ANALYTICS & O.R. APPLICATIONS
For submission information:
http://meetings.informs.org/analytics2013
Big Data
Marketing Analytics
Forecasting
Soft Skills for Analysts
Supply Chain Management
The Analytics Process
Decision Analysis
Analytics for HR
Software Tutorials
Thanks to our Sponsors
Choose
from these
Focused Tracks
Your registration fee admits you to the gala Edelman Award Banquet on
Monday evening. Come to salute the Edelman Award winner, as well as the
winners of the Wagner Prize, INFORMS Prize, UPS George D. Smith Prize and
Innovative Applications in Analytics Award.
Analytics Analytics 2pg. ad option B_Layout 1 2/25/13 5:04 PM Page 1
April 7-9, 2013
Grand Hyatt
San Antonio,
Texas
Showcasing
THE POWER
OF
ANALYTICS
BOEING
IBM
TARGET
GOOGLE
MIT
INTEL
DELL
USAA
GARTNER
STARBUCKS
FORD
NEILSEN
STANFORD
BOOZ ALLEN
HAMILTON
CHEVRON
SAS
PEPSICO
TERADATA
GE
KROGER
DELOITTE
and 70 other top companies,
universities and government agencies
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.
• Real-world use of descriptive, PREDICTIVE and prescriptive analytics
• Focus on BIG DATA, marketing analytics, forecasting
• Most rigorous and REAL-WORLD analytics conference offered
• Administration of Certified Analytics Professional exam
• Not just what to do but HOW TO DO IT
Keynote Speakers
Edelman Award Ceremony & Banquet
Nicole Piasecki
Vice President,
Business Development
& Strategic Integration
Boeing Commercial Air-
planes
Sandy Carter
Vice President
IBM Social Business
Evangelism
and Sales
IBM
Learn from the BEST about High-Impact
ANALYTICS & O.R. APPLICATIONS
For submission information:
http://meetings.informs.org/analytics2013
Big Data
Marketing Analytics
Forecasting
Soft Skills for Analysts
Supply Chain Management
The Analytics Process
Decision Analysis
Analytics for HR
Software Tutorials
Thanks to our Sponsors
Choose
from these
Focused Tracks
Your registration fee admits you to the gala Edelman Award Banquet on
Monday evening. Come to salute the Edelman Award winner, as well as the
winners of the Wagner Prize, INFORMS Prize, UPS George D. Smith Prize and
Innovative Applications in Analytics Award.
Analytics Analytics 2pg. ad option B_Layout 1 2/25/13 5:04 PM Page 1
WWW. I NF OR MS . OR G 42 | A NA LY T I CS - MAGA Z I NE . OR G
The Dow Chemical approach to leveraging time-series
data and demand sensing.
Big data means different
things to different people. In
the context of forecasting,
the savvy decision-maker
needs to fnd ways to derive value from
big data. Data mining for forecasting of-
fers the opportunity to leverage the nu-
merous sources of time series data, both
internal and external, now readily avail-
able to the business decision-maker, into
actionable strategies that can directly im-
pact proftability. Deciding what to make,
when to make it and for whom is a com-
plex process. Understanding what fac-
tors drive demand, and how these factors
(e.g., raw materials, logistics, labor, etc.)
Integrating
data mining and
forecasting
BY TIM REY (LEFT) AND CHIP WELLS
B
CASE STUDY
MAR CH / A P R I L 2013 | 43 A NA L Y T I C S
interact with production processes or de-
mand and change over time are keys to
deriving value in this context.
The Dow Chemical Company was in-
terested in developing an approach for
demand sensing that would provide:
Cost reduction
• reduction in resource expenses for
data collection and presentation
• consistent automated source of data
for leading indicator trends
Agility in the Market
• shifting to external and future looks
from internal history
• broader dissemination of key leading
indicator data
• better timing on market trends
… faster price responses, better
resource planning (by reducing
allocation/force major/share loss on
the up side and reducing inventory
carrying costs and asset costs on the
down side)
Improved accuracy
• accuracy of timing and estimates for
forecast models
Visualization
• understanding leading indicator
relationships
Dow (and its Advanced Analytics
team) was keenly interested in better
forecasting models for volume (demand),
net sales, standard margin, inven-
tory costs, asset utilization and EBIT
Figure 1: Levels of hierarchy at Dow Chemical.
WWW. I NF OR MS . OR G 44 | A NA LY T I CS - MAGA Z I NE . OR G
(earnings before interest and taxes). This
was to be done for all businesses and all
geographies. Similar to many large cor-
porations, Dow has a complex business/
product hierarchy. This hierarchy starts
at the top, total Dow, then moves down
through divisions, business groups,
global business units, value centers, per-
formance centers, etc. As is the case in
DATA & DEMAND SENSI NG
Figure 2: Dow’s value chains are deep and complex.
Figure 3: Target variables of interest are generally related to one another.
most large corporations, this hierarchy
is always changing and is overlaid with
geography. Even lower levels of the hi-
erarchy exist when specifc products are
considered.
Dow operates in the vast majority
of the 16 global market segments as
defined in the ISIC (International Stan-
dard Industrial Classification) market
MAR CH / A P R I L 2013 | 45
A NA L Y T I C S
segment structure, some of which
are: agriculture, hunting and forestry,
mining and quarrying, manufacturing,
electricity, gas and water supply, con-
struction, wholesale and retail trade,
hotels and restaurants, transport, stor-
age and communications, health and
social work, etc. This includes com-
modities, differentiated commodities
and specialty products and thus makes
the mix even more complex. The val-
ue chains Dow is involved in are very
deep and complex, and often connect
the earliest stages of hydrocarbons
extraction and production all the way
to the consumer on the street.
Before embarking on the project,
the team contemplated a few “indus-
trial” and economic considerations to
attack. First, simply multiplying out the
number of models, the team saw that
they would have around 7,000 exog-
enous variable models to build, so we
focused on the top global business
units (by area combinations in each
division, restricting our initial effort to
covering 80 percent of net sales). Next,
we realized that the target variables
of interest (volume, asset utilization,
net sales, standard margin, inventory
costs and EBIT) are generally related
to one another. Thus, volume is a func-
tion of volume “drivers” (Vx), repre-
sented by f(Vx); asset utilization (AU)
is a function of volume and AU “driv-
ers” f(AUx); inventory is a function of
volume and inventory (INV) “drivers”
f(INVx); net sales is driven by volume,
various costs (xcosts) and net sales
“drivers” f(NSx); standard margin is
driven by net sales and standard mar-
gin “drivers” f(SMx); and finally EBIT
is driven by standard margin and EBIT
“drivers” f(EBITx).
The problem, if done only at one level
of the hierarchy, fts into a multivariate in
Y approach that could be solved using a
VARMAX (vector auto regressive mov-
ing average with exogenous variables)
system. The complexity here is that we
needed to solve the problem across the
hierarchy shown above. We proposed
that we could mimic the VARMAX struc-
ture by building the models in a “daisy
chain” fashion shown in Figure 3. As a
baseline, we thus compared a traditional
Join the Analytics Section of INFORMS
For more information, visit:
http://www.informs.org/Community/Analytics/Membership
WWW. I NF OR MS . OR G 46 | A NA LY T I CS - MAGA Z I NE . OR G
VARMAX approach to the daisy chain approach at
the total Dow level. We also did a traditional univari-
ate model, as well as a traditional ARIMAX model for
each Y. The “Reconciled” column in Table 1 was the
daisy chain approach used in the hierarchy (imple-
mented via SAS Forecast Studio) and then reconciled
up. Given the results in Table 1, we were confdent
we could use the daisy chain approach across the
hierarchy and get similar beneft to the VARMAX ap-
proach. All of the above was accomplished with vari-
ous SAS forecasting platforms.
Following the data mining for forecasting process
described in “Applied Data Mining For Forecasting Us-
ing SAS” (Rey, Kordon and Wells (2012)) – Chapters 2
and then 7 – which covers exogenous variable identif-
cation and then Reduction and Selection for forecasting
leads to conducting dozens of mind mapping sessions
to have the businesses propose various sets of “driv-
ers” for the numerous GBU and VC by geographic area
combinations. This leads to using thousands (more than
15,000 in this case!) of potential exogenous variables
DATA & DEMAND SENSI NG
Table 1: SAS Forecast Studio screen shot.
Request a no-obligation INFORMS Member Benefits Packet
For more information, visit: http://www.informs.org/Membership
MAR CH / A P R I L 2013 | 47
A NA L Y T I C S
of interest for the 7,000 models in the hier-
archy. This is truly a big data, large-scale
forecasting problem. A lot of automation
was necessary for frst setting up initial re-
search projects, as well as automatically
building initial univariate and daisy chain
models.
Lastly, concerning visualization, the
business can gain access to these fore-
casts in a corporate-wide business intel-
ligence delivery system where they can
see the history, model, forecast, conf-
dence limits and drivers.
Big data mandates big judgment.
Big judgment has to have short “ask-
to-answer” cycles. These opportuni-
ties call for the use of data mining for
forecasting approaches that lead to
using special techniques for variable
reduction and selection on time series
data. ❙
Tim Rey ([email protected]) is director of Advanced
Analytics at The Dow Chemical Company. Fenton
(Chip) Wells ([email protected]) is a statistical
services specialist in SAS Education at SAS. They
are co-authors of the book, “Applied Data Mining
and Forecasting Using SAS.”
Nomination Requirements and Process
http://meetings.informs.org/Analytics2013/profcolloquium.html
PROFESSIONAL COLLOQUIUM Sunday, April 7, 2013
Real-World Career Guidance for Practice-Oriented Graduate Students
This interactive colloquium is designed to
help practice-oriented master’s and PhD
students transition into successful real-world
careers. This colloquium presents the
experience & perspective of practitioners
who have bridged the gap in order to
give attendees a jump-start in their
new careers.
IPC registration fee of $375 covers
all these benefits:
· Colloquium on Sunday, April 7
· Full registration to the INFORMS
Business Analytics and Operations
Research Conference, April 7-9
(includes all meals)
http://meetings.informs.org/analytics2013/profcolloquium.html
Nomination deadline:
March 11
WWW. I NF OR MS . OR G 48 | A NA LY T I CS - MAGA Z I NE . OR G
Statistical
software in the
Age of the Geek
BY JAMES J. SWAIN
SOFTWARE SURVEY
MAR CH / A P R I L 2013 | 49 A NA L Y T I C S
We are witnessing the emer-
gence of a data-rich culture
that has been made possible
by intelligent devices, online
commerce, social networks and scien-
tifc sensors – all ceaselessly creating,
exchanging and collecting data. There is
so much data that we have had to coin
new words to describe the quantities –
terabytes, petabytes, exabytes – just
to keep up. What secrets can we divine
from within this Everest of data, whether it
leads to a responsive global supply chain
or window into the souls of consumers?
There is a need to make sense of it all,
so perhaps it is no surprise that Google’s
chief economist Hal Varian declares that
statisticians surely have the “really sexy
job” for the coming decade, for who better
will there be to make sense of so much
data? What is equally certain is that soft-
ware tools will be needed to process and
analyze the data.
ELECTION POLLS AND PREDICTIONS
The recent presidential election pro-
vides a microcosm of the phenomena.
Due to such intense interest in elections,
even statistics become interesting when
they tell a story about the winners and
the losers. Opinion polling began early
last century and is constantly being re-
fned in technique and frequency; dur-
ing this last election multiple polls were
made and updated throughout the elec-
tion cycle. Election predictions using a
computer and early election results are
now 60 years old, dating to a UNIVAC
computer that predicted a landslide for
Ike in 1952. The technology and analysis
has grown in sophistication. Even better,
media has improved its ability to animate
a story of progress or retreat in the quest
of congressional seats or electoral votes.
The experts, buttressed by numbers,
confdently assessed the “bounce” of the
party convention, the “loss” of a poor de-
bate performance or the impact of the lat-
est economic news.
One indication that polling results
were considered important to the per-
ception and eventual results of the cam-
paigns was illustrated by controversies
over both opinion polls and key govern-
mental statistics such as the unemploy-
ment rate. Nate Silver of The New York
Times was maligned for early predicting
an Obama win in the election using a
model that aggregated the results from a
number of polls. Polling using small sam-
ples in volatile election is always tricky,
W
Survey Directory & Data
To view the directory of statistical
software products along with the survey
data, click here.
WWW. I NF OR MS . OR G 50 | A NA LY T I CS - MAGA Z I NE . OR G
as opinions can shift in the time that it takes to ob-
tain a complete sample, though these uncertainties
should average out when combined.
Of course, we have come a long way since the Lit-
erary Digest fasco of 1936, which predicted a land-
slide for Alf Landon in the presidential election based
on a badly fawed sample of 2.4 million respondents.
In the event, Landon only carried two states. In that
same year George Gallup sampled only 50,000 re-
spondents but was able to accurately predict the
results for Franklin D. Roosevelt. Typical polls now
take samples of only 800 to 1,200 likely voters, with
pains taken to ensure that sampling is random and
representative of the general population (e.g., likely
voters) of interest. During the election anything that
might infuence public opinion was questioned. This
included government statistics such as the decline
in the unemployment rate announced by Bureau of
Labor Statistics just before the election.
On election night, we were treated to a deluge
of results in battleground states such as Ohio,
Pennsylvania, Virginia and Florida, both visually
and numerically. Each state had its own chorop-
leth map, with results summarized by county or
precinct, with results dissected according to vari-
ous demographic groupings. The great unknown
through the evening was the number of voters in
various categories and their turn out, since the
relative rate at which these groups supported in-
dividual candidates was better estimated.
While the presentation of data was impressive
during the election, perhaps the most interesting
story was the highly effective use that the Obama
campaign made of analytics to fne-tune donation
STATI STI CAL SOFTWARE SURVEY
While the presentation
of data was impressive
during the election,
perhaps the most
interesting story was the
highly effective use that
the Obama campaign
made of analytics to fine-
tune donation pitches,
simplify the process of
contacting and supporting
the campaign, make media
buys and coordinate field
organizers.
SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. ® indicates USA registration. Other brand and product names are trademarks of their respective companies. © 2013 SAS Institute Inc.
All rights reserved. S105652US.0213
sas.com/know
Learn why Forrester Research Inc. named SAS “an analytics powerhouse.”
(Forrester Research, Inc., The Forrester Wave

:
Big Data Predictive Analytics Solutions, Q1 2013)
Know what’s hot.
The topic of analytics is on fre right now. With SAS,
®
you can discover innovative ways to increase profts,
reduce risk, predict trends and turn data assets — whether big or small — into competitive advantage.
Decide with confdence.
WWW. I NF OR MS . OR G 52 | A NA LY T I CS - MAGA Z I NE . OR G
pitches, simplify the process of contact-
ing and supporting the campaign, make
media buys and coordinate feld orga-
nizers. As documented in the 90-page
report, “Inside the Cave,” the technol-
ogy team for Obama was signifcantly
larger than for the Romney campaign,
had more than four times the number
of donors and fve times the number of
e-mails addresses. His staff was drawn
from technologists, volunteers who were
recruited directly from technology hot-
beds such as the Silicon Valley (including
Rayid Ghani, former director of analyt-
ics research at Accenture). The Obama
campaign raised more than $690 million
online, increasing both the number of in-
dividuals and the average donation com-
pared to 2008. Throughout the campaign
the Obama campaign experimented with
a variety of messages, monitored the
results and fne-tuned their pitches and
even the subject line of e-mails to make
them as effective as possible. The cam-
paign hired exclusively for technology
skills and gave them free reign to learn
from the data with “little to no interfer-
ence from campaign management on
STATI STI CAL SOFTWARE SURVEY
Request a no-obligation INFORMS Member Benefits Packet
For more information, visit: http://www.informs.org/Membership
content.” The team also performed ex-
tensive surveys and tracking polls in key
states, increasing in frequency as the
election neared. The team used the data
collected to build models and performed
simulations to gauge progress and to al-
locate resources dynamically.
During the presidential campaign,
the news media thrived on colorful and
dynamic visual representations of the
data. The tools in this statistical soft-
ware survey provide a variety of ways
to present data to make comparisons,
demonstrate trends or search for outli-
ers within the data. Further tools and
some guidelines for good graphical
design are provided in “Visual This”
(Yau, 2011). The author has a website
FlowingData (http://flowingdata.com/)
that provides further examples. The
website GapMinder (http://www.gap-
minder.org/) provides examples of ani-
mated graphics that offer methods of
exploring multiple variables over time
among different regions of the world.
The stated aim of the website is to pro-
vide information about the world (e.g.
life expectancy, income, birthrates) to
MAR CH / A P R I L 2013 | 53
A NA L Y T I C S
facilitate discussion. The use of ani-
mation provides an additional way to
perceive relations among and between
multivariable factors as a function of
another variable, such as time.
DATA
The computer and Internet have
revolutionized the acquisition, storage
and access to data in addition to provid-
ing the computation power to numeri-
cally process the data. Public and private
sources exist for a wide range of data,
including demographics, production, in-
vestment, monetary, infection, disease,
mortality and all manner of consumer
and operational data. This has given rise
to new techniques in multivariate dimen-
sion reduction, data mining and machine
learning.
Join the Community for inside information.
SECTION ON
ANALYTICS
2013 Innovative Applications in Analytics Award
Finalist Presentations Now Online
http://analytics.section.informs.org
Organizations want to take
advantage of tools and techniques
to improve their decision making
but don’t always know how to start.
View these on-demand videos from
top-notch analytics experts to
improve your analytics maturity.
Help Promote Analytics
It’s fast and it’s easy! Visit:
http://analytics.informs.org/button.html
WWW. I NF OR MS . OR G 54 | A NA LY T I CS - MAGA Z I NE . OR G
The computer has made it possible to expand
the range of what is meant by data, from purely nu-
merical data to text and graphics. Documents and e-
mails can be compared using word frequencies and
even stochastic models, and user choices online can
be logged and analyzed for patterns. Prediction of
user choices is useful both to accelerate searching,
as well as to match advertising to users. To obtain
an idea of the magnitude of data available, consider
the data provided in the recent Netfix prize competi-
tion. The objective was to predict user movie ratings
based on a training data set of more than 100 million
ratings from over 480,000 users. The qualifying data
used to evaluate the proposed algorithms consisted
of more than 2.8 million user and movie combina-
tions whose ratings had to be predicted. Of course,
even these huge samples are but a fraction of the
data collected on the Web.
STATISTICAL SOFTWARE SURVEY
The 2013 survey of statistical software pro-
vides capsule information about various products
and vendors. The tools range from general tools
that cover the important techniques of inference
and estimation, as well as specialized activities
such as nonlinear regression, forecasting and de-
sign of experiments. The product information con-
tained in the survey was obtained from product
vendors and is summarized in the following tables
to highlight general features, capabilities, comput-
ing requirements and to provide contact informa-
tion. Many of the vendors have their own websites
for further, detailed information, and many provide
demonstration programs that can be downloaded
STATI STI CAL SOFTWARE SURVEY
The computer has made
it possible to expand the
range of what is meant
by data, from purely
numerical data to text and
graphics. Documents and
e-mails can be compared
using word frequencies
and even stochastic
models, and user choices
online can be logged and
analyzed for patterns.
MAR CH / A P R I L 2013 | 55
A NA L Y T I C S
from these sites. No attempt is made
to evaluate or rank the products, and
the information provided comes from
the vendors themselves. To view the
survey data, click here. Vendors that
were unable to make the publishing
deadline will be added.
Products that provide statistical add-
ins available for use with spreadsheets
remain popular and provide enhanced
specialized capabilities for spread-
sheets. The spreadsheet is the primary
computational tool in a wide variety of
settings, familiar and accessible to all.
Many procedures of data summarization,
estimation, inference, basic graphics and
even regression modeling can be added
to spreadsheets in this way. An example
is the Unistat add-in for Excel. The func-
tionality of products for use with spread-
sheets continues to grow, including risk
analysis and Monte Carlo sampling, such
as Oracle Crystal Ball.
Dedicated general and special pur-
pose statistical software generally have
a wider variety and depth of analysis
than available in the add-in software. For
many specialized techniques such as
forecasting, design of experiments and
so forth, a statistical package is appropri-
ate. In general, statistical software plays
a distinct role on the analyst’s desktop,
and provided that data can be freely ex-
changed among applications, each part
of an analysis can be made with the most
appropriate (or convenient) software tool.
An important feature of statistical pro-
grams is the importation of data from as
many sources as possible to eliminate
the need for data entry when data is al-
ready available from another source. Most
programs have the ability to read from
spreadsheets and selected data storage
formats. Also highly visible in this survey is
the growth of data warehousing and “data
mining” capabilities, programs and training.
Data mining tools attempt to integrate and
analyze data from a variety of sources (and
purposes) to look for relations that would
not be possible from the individual data
sets. Within the survey we observe several
specialized products, such as STAT::FIT,
which is more narrowly focused on distri-
bution ftting than general statistics, but of
particular use to developers of stochastic
models and simulation. ❙
James J. Swain ([email protected]) is
professor and chair, Department of Industrial and
Systems and Engineering Management, at the
University of Alabama in Huntsville. He is a senior
member of INFORMS. He is also a member of
ASA, IIE, and ASEE.
REFERENCES
1. “Inside the Cave: An In-depth look at the
digital, technology, and analytics operations of
Obama for America,” Engage Research,
www.engagedc.com/inside-the-cave/.
2. Nathan Yau, 2011, “Visualize This: The
FlowingData Guide to Design, Visualization and
Statistics,” Wiley, New York, 2011.
WWW. I NF OR MS . OR G 56 | A NA LY T I CS - MAGA Z I NE . OR G
CORPORATE PROFI LE
In November 1986, the Man-
agement Science Group
was established at Merrill
Lynch. The initial group was
comprised of six operations research/
management science practitioners who
previously worked together at the RCA
Operations Research Group, which was
founded by Franz Edelman circa 1950.
DuWayne Peterson, head of Opera-
tions, Systems and Telecommunications
at Merrill Lynch in 1986, was the driving
force behind the idea to build an analytics
team within the brokerage business. He
believed that an internal team of man-
agement scientists could identify a host
of opportunities to improve business per-
formance. Since RCA had recently been
acquired by General Electric, the RCA
O.R. group was looking for a new chal-
lenge as well as a new home. Over the
past 25 years, Peterson’s decision paid
off with big dividends.
The mission of the group was unique
in fnancial services. There were and still
are numerous analysts on Wall Street fo-
cused on developing investment strate-
gies, managing risk, supporting trading
25 years of
analytics at
Bank of America
BY RUSS LABE
I
MAR CH / A P R I L 2013 | 57
A NA L Y T I C S
activity and seeking arbitrage
opportunities. Our idea was to
focus instead on supporting the
general business functions of
the organization, to improve ef-
fciency and apply analytics to
key business functions such as
operations, marketing, pricing
and product management.
Much has happened since
1986. Executive leadership
changed several times. Finan-
cial products and solutions
have evolved, grown and declined. Bank
of America acquired Merrill Lynch in
2009. The size of the group grew, shrank
and grew again. The name of the group
changed to Decision Support Modeling.
The group migrated to different functional
areas of the organization, including tech-
nology, marketing, strategic planning and
the brokerage line of business. Three
different leaders managed the team. H.
Newton Garber, the original director, re-
tired in 1990. Garber was followed by
Raj Nigam, who led the group until he
retired in 2004. Today, the group is man-
aged by Russ Labe, one of the original
founding members, and includes 10 pro-
fessionals with more than 120 years of
combined experience at Bank of Amer-
ica and Merrill Lynch. It is part of the
marketing organization supporting the
client-managed businesses within Bank
of America, which includes Merrill Lynch
Wealth Management, U.S. Trust and the
Bank of America Merrill Lynch institution-
al businesses.
Throughout all those changes, the
group has stayed focused on its mission
to improve proftability and assess stra-
tegic decisions by providing statistical
analysis and mathematical modeling.
APPLICATION AREAS
While the focus of the team’s project
work has certainly evolved over time, certain
application areas – pricing, client retention,
product propensity and revenue forecasting
– are consistently important to the business
and provide recurring projects.
The group has evaluated numerous
pricing situations related to new products
and restructuring the pricing of existing prod-
ucts and solutions. Typically these models
A view of the Bank of America Merrill Lynch corporate
campus in Hopewell, N.J.
WWW. I NF OR MS . OR G 58 | A NA LY T I CS - MAGA Z I NE . OR G
required gathering huge amounts of trans-
actional data and building historical simula-
tions to evaluate alternate pricing scenarios.
The analysis was focused on understanding
the detailed impact of price changes on indi-
vidual clients, fnancial advisor compensa-
tion and frm proftability. In phase one of the
analysis, the team assumed client behavior
remains the same in order to establish an
initial estimate of change impact. In phase
two of the analysis the team simulated
changes in client behavior based on factors
such as client satisfaction, fnancial advisor
loyalty and transactional behavior patterns.
Sometimes the resulting analysis led to a
decision to not implement any changes.
In other cases, the analysis helped launch
new products, solutions and services.
Models the team developed to identify
clients at risk of leaving Merrill Lynch are
used on a regular basis. Alerts are distrib-
uted to branch offces each week so they
can save some of those relationships. This
program has been in place for more than 10
years and is estimated to save $1 billion of
client assets annually. The underlying mod-
els were developed using a combination
of decision trees and logistic regression. A
signifcant refresh and update of the models
was recently completed and an enhanced
program implementation is in progress.
Over the years, the success of this work
CORPORATE PROFI LE
Members of the Decision Support Modeling team (l-r): Lihua Yang, Mark Goldstein, Fang Liu, Je
Oh, Zhaoping Wang, Yanni Papadakis, Vera Helman, Russ Labe and Brian Jiang.
MAR CH / A P R I L 2013 | 59
A NA L Y T I C S
led to related applications, including devel-
opment of customized models specifcally
focused on commercial or small business
clients and Merrill Edge clients. Merrill Edge
consists of the Merrill Edge Advisory Center
(a call center providing investment guidance
to clients), as well as self-directed online
investing.
The team developed numerous product
propensity models to help target the most
appropriate fnancial solutions and services
for the most appropriate clients. This has
been a highly active area for many years.
Originally the team developed customized
models for each product and solution of
interest. Over the last few years the team
developed an automated process, based
on collaborative fltering, that allows us to
automatically update more than 70 product
propensity models and score more than two
million clients across all the models on a
monthly basis. The models are customized
by client segment, resulting in more than
1,000 separate models that are revised
each month. These models are used as
input to various marketing campaigns and
client contact optimization strategies and
models.
A related application is referral models,
a type of propensity model that addresses
a strategic priority for the bank – identify op-
portunities to support clients across multiple
lines of business. Examples include identify-
ing consumer banking clients with a need
to manage their investment accounts and
fnding small business owners with a need
to manage their personal investments. The
team developed a series of these referral
models, which are used to help custom-
er service representatives provide better
service to clients and to inform marketing
campaigns by selecting the most appro-
priate messages for each client. Results
from these models are embedded in the
bank’s Web sites to determine messaging
strategies.
Another application area is revenue
forecasting models. Recently, this has
become an area of heightened interest
to meet new regulatory requirements
imposed by the Federal Reserve. The
models estimate the impact of economic
stress scenarios, defned quarterly by
regulators, on the bank’s fnancial per-
formance. The team developed a series
of econometric regression models that
predict monthly revenue over a two-year
horizon based on a combination of mac-
ro-economic factors, internal business
drivers and seasonality. The models are
used each quarter to support the required
stress test analysis. In addition, Finance
uses the models as input to their ongoing
planning process. These models were
classifed as trade secrets by the bank.
Other examples of the team’s work
include fnancial advisor segmentation,
client proftability models, advertising
WWW. I NF OR MS . OR G 6 0 | A NA LY T I CS - MAGA Z I NE . OR G
impact modeling and measurement, f-
nancial advisor compensation analysis,
business strategy impact evaluations, re-
serve requirement models for debit card
reward points and deferred compensa-
tion programs, and portfolio optimization.
Benefts from these projects through the
years have impacted strategic business
decisions and contributed hundreds of
millions of dollars in bottom line benefts
through increased revenue, cost reduc-
tion and effciency improvements.
PROFESSIONAL RECOGNITION
In addition to internal contributions,
the team has received external rec-
ognition from INFORMS for the qual-
ity and business value of its work. In
1997, the team helped Merrill Lynch
win the INFORMS Prize for the effec-
tive and widespread use of analytics
throughout the organization sustained
over 10 years.
In 2001, the team helped Merrill
Lynch win the Franz Edelman Prize for
the pricing analysis it conducted to help
launch Merrill Lynch Unlimited Advan-
tage (MLUA), as well as the legacy ML
Direct business, now called Merrill Edge.
MLUA was the frst fnancial solution in
the brokerage marketplace with true cli-
ent relationship pricing and attracted $22
billion of incremental assets to the frm
during its frst two years. ML Direct was
the frm’s frst effort in Web-based online
trading accounts.
In 2004, the team won the Wagner
Prize for modeling the liquidity risk of re-
volving credit lines provided to other com-
panies through Merrill Lynch’s legacy Bank
and Trust business. This analysis allowed
ML Bank and Trust to free up $4 billion of
capital. In 2005, the team helped the ML
Treasury group win the Alexander Hamil-
ton Prize associated with the same work
around liquidity risk. These awards from ob-
jective, professional peers helped raise the
profle of the group internally and increased
management confdence in the quality and
value of the team’s work.
SUMMARY
Looking back over the last 25 years, the
Decision Support Modeling team at Bank of
America has much to celebrate. The team
survived many market cycles and reorga-
nizations. The team has a strong history of
business impact on the organization, using
advanced analytics to support good man-
agement practices and business transfor-
mation. The group provided contributions
on a wide range of business issues, includ-
ing pricing, client attrition, product propensi-
ty, fnancial advisor segmentation, revenue
forecasting, fnancial advisor compensation,
business strategy impact evaluations and
portfolio optimization. The team employed a
wide range of modeling techniques across
CORPORATE PROFI LE
MAR CH / A P R I L 2013 | 61
A NA L Y T I C S
Global Wealth & Investment Management is a division of
Bank of America Corporation (“BAC”). Merrill Lynch Wealth
Management, Merrill Edge™, U.S. Trust and Bank of America
Merrill Lynch are affliated sub-divisions within Global Wealth
& Investment Management.
Merrill Lynch Wealth Management makes available products
and services offered by Merrill Lynch, Pierce, Fenner & Smith
Incorporated (“MLPF&S”) and other subsidiaries of BAC. Mer-
rill Edge™ is the marketing name for two businesses: Merrill
Edge Advisory Center, which offers team-based advice and
guidance brokerage services; and a self-directed online in-
vesting platform.
U.S. Trust, Bank of America Private Wealth Management op-
erates through Bank of America, N.A., and other subsidiaries
of BAC.
Bank of America Merrill Lynch is a marketing name for the
Retirement & Philanthropic Services businesses of BAC.
Banking products are provided by Bank of America, N.A., and
affliated banks, Members FDIC and wholly owned subsidiar-
ies of BAC.
Investment products are not FDIC insured, are not bank guar-
anteed and may lose value.
different types of projects, including data
mining, design of experiments, multi-vari-
ate statistics, simulation, and optimization.
Looking to the future, the team sees many
opportunities to continue applying analyt-
ics and modeling at Bank of America and to
continue providing added value to the busi-
ness. ❙
Russ Labe ([email protected]) is
director of the Decision Support Modeling Group,
Bank of America Corporation. He is a member of
INFORMS Roundtable and a senior member of
INFORMS.
Baosteel, for "Operations Research Transforms Baosteel’s Operations"

Chevron, for "Optimizing Chevron’s Refineries"

Dell, for "Dell’s Retail Transformation through Analytics"

Delta Commissioner of Holland, for "Economically Efficient Flood Standards to
Protect the Netherlands against Flooding"

Kroger, for "Simulation and Optimization Improves Pharmacy Inventory Management
at the Kroger Co."

McKesson, for "A Holistic Supply Chain Sustainability Management Solution"
http://meetings.informs.org/Analytics2013
Be there at the Edelman Gala, April 8 when the 2013 winner is announced!
WWW. I NF OR MS . OR G 62 | A NA LY T I CS - MAGA Z I NE . OR G
Sandy Carter, VP of Social Business
Evangelism and Sales at IBM, will deliver
the keynote presentation at the 2013 IN-
FORMS Conference on Business Analyt-
ics and O.R., April 7-9, in San Antonio,
Texas. Carter¹s thought-provoking talk
will highlight three days of intensive, real-
world education on descriptive, predictive
and prescriptive analytics.
The program will feature presentations
by more than 100 speakers representing
a broad range of industries and applica-
tions areas (see sidebar story on page
64). Two expanded, 10-session tracks
will focus on marketing analytics and the
analytics process. The marketing track
includes talks by executives from Target,
Nielsen, Mars, USAA, Emory University
and other leading organizations, covering
everything from marketing mix optimiza-
tion to stochastic models for customer an-
alytics. The analytics process track offers
sessions on topics such as distribution
processing and hospital business analyt-
ics, presented by speakers from Stanford,
GE, Schneider National, SAS and Strata
Decision Technology.
INFORMS Analytics
Conference
Focus on marketing analytics, big data, HR analytics and much more.
CONFERENCE PREVI EW
New this year are focused talks on
big data, as well as a track on analyt-
ics for human resources management.
Speakers from Ford, Teradata, IBM
and the U.S. Internal Revenue Service
will discuss both the technological and
business aspects of big data. The HR
track features speakers from Google,
PepsiCo, IBM and Abbott Consulting.
Other tracks will cover supply chain
management, forecasting, decision
analysis and soft skills for analysts.
The Franz Edelman Competition,
showcasing the best in high-impact ana-
lytics and O.R., will feature presentations
by six fnalists: Baosteel, Chevron, Dell,
Delta Commissioner of Holland, Kroger
and McKesson. A new track for 2013 will
highlight applied work that has received
recognition in other competitions, includ-
ing fnalists for the Innovative Applica-
tions in Analytics Award and winners of
the Wagner Prize, INFORMS Prize, UPS
George D. Smith Prize and Spreadsheet
Guru Prize.
For conference information: http://
meetings2.informs.org/Analytics2013 ❙
Analytics is the scientific process of transforming data into insight for making better decisions. Our
members help organizations like yours everyday use analytics to improve processes, save costs, and
enhance revenues.
WANT TO KNOW MORE?
http://www.informs.org/Sites/Getting-Started-With-Analytics
Our Getting Started with Analytics website provides:
• Over 80 searchable “quick-bite” success stories
• Five signs you can benefit from analytics
• Information on how to recognize the opportunity
• Sources of analytics expertise
• What to expect when working with an analytics professional
Start Here and Contact us for more information.
Find out how the use of analytics can improve your bottom line.
INFORMS can help!
Get Started with Analytics
No matter your organizational situation, INFORMS and the skilled use of analytics can help.
Getting Started With Analytics
Analytics in the News
MORE ANALYTICS SUCCESS STORIES Industry Function Benefit
Search Analytics Portal
WHAT’S YOUR TOUGHEST, MOST
COMPLEX PROBLEM?
START HERE »
Home Start Here What Analtyics Is What it Can Do For You How To Start Using Analytics More Resources The Edelman Award Analytics Success Stories
WWW. I NF OR MS . OR G 64 | A NA LY T I CS - MAGA Z I NE . OR G
CONFERENCE PREVI EW
MARKETING ANALYTICS
• Target: Maher Lahmar, Manager R&D, Merchandising
Business Intelligence, and Rob Parkin, Manager-
Group Lead, IBM, on assortment planning
• Nielsen Marketing Analytics: Nathan Brixius, Sr. VP of
Analytics Development, on marketing ROI analytics –
from data to insight
• Emory University: David Schweidel, Associate Professor,
Co-Director, Emory Marketing Analytics Center, on stochastic
models for customer analytics
• USAA: Allen S. Crane, Executive Director, Applied
Analytics, on cross-channel insights – seeing the big
picture to create ultimate member experiences
• Mars: Sorin Patilinet, Marketing Services Manager,
on analytics to measure the direct impact of TV
advertising on sales
• PROS: Neil Biehn, VP of Science and Research, on
the new “V” in big data – how viability is driving better
outcomes
• Baylor University: John Tanner Jr., Professor of
Marketing, Associate Dean, on illuminating the path to
purchase
• ThinkVine: Damon Ragusa, Chairman & Chief
Strategy Offcer, on using agent-based modeling to
power marketing mix optimization
BIG DATA
• U.S. Internal Revenue Service: Jeff Butler, Director,
Research Databases, on performance strategies for
analyzing big data
• Ford: Michael Cavaretta, Technical Leader, Predictive
Analytics, on where will big data and data science
take O.R.?
• IBM Research: Amol N. Ghoting, Research Staff
Member, on implementing big data analytics
• Teradata: Priyank Patel, Director of Product Management,
on multi-genre discovery analytics on big data
ANALYTICS PROCESS
• General Electric-Power & Water: Sameer Vittal, Manager
of Advanced Analytics, on analytics in wind energy
• SAS: Kathy Lange, Senior Director, Americas
Business Analytics Practice, on the effective use of
business analytics
• Breeden Ideas: Jake Breeden, Principal, on stopping
sacred cows before they stop analytics
• Schneider National: Ted L. Gifford, Distinguished
Member of Technical Staff, on integrated analytics in
transportation and logistics
• Stanford University: Sam L. Savage, Consulting
Professor, Management Science & Engineering, on
distribution processing, a cure for the faw of averages
• Booz Allen Hamilton: Juergen A. Klenk, Principal, on
generating value from data
• Strata DecisionTechnology: Don N. Kleinmuntz,
Executive VP and Chief Analytics Offcer, on analytics
opportunities and challenges in healthcare
• Princeton Consultants: Steve Sashihara, President &
CEO, on optimizing as if people matter
• Ghent University” Dirk Van den Poel, Professor of
Marketing Analytics, on training analysts for business
SUPPLY CHAIN MANAGEMENT
• Gartner: Stan Aronow, Supply Chain Research Director,
on maximizing proftability by integrating cost-to-serve
and supply chain optimization
• MIT: Chris Caplice, Executive Director, MIT Center for
Transportation & Logistics, on optimal allocation of for-
hire and private feets
• Supply Chain Insights: Lora Cecere, CEO, on supply
chain technology: Where are we headed?
• Intel: Thomas Rucker, Director of Technology
Development, on concurrent multi-phase development
to ensure a vibrant supply chain
• Sedlak Management Consultants: George Swartz,
Jr., Practice Director, on distribution network analysis
& design, helping Nigerians gain access to HIV/AIDS
treatment
DECISION ANALYSIS
• Decision Strategies: William Haskett, Senior Principal,
on strategy arenas, a path to objective objectives and
superior strategies
• Lumina Decision Systems: Max Henrion, CEO, on
generating real business value from data
• Innovative Decisions: Freeman Marvin, Vice
President, on the VII soft O.R. methods every analyst
must know
Focused Tracks: Inside Look at Analytics Best Practices
MAR CH / A P R I L 2013 | 65
A NA L Y T I C S
• U.S. Army: Col. Brian K. Sperling, Senior Military
Advisor/Military Deputy Director, Center for Army
Analysis, on ORSA … embedded in Army operations
FORECASTING
• SAS: Michael Gilliland, Product Marketing Manager,
on process control approaches in business
forecasting
• Arizona State University: Dieter Armbruster, Professor,
Mathematical & Statistical Sciences, on characterizing
and forecasting the market infuence of OEMs
• Planalytics: Scott A. Bernhardt, President, on measuring
and managing the weather’s impact on business
• SCM Focus: Shaun Snapp, Consultant, Founder &
Editor, on supply-planning friendly demand planning
SOFT SKILLS FOR ANALYSTS
• Chevron: William K. Klimack, Decision Analysis
Consultant, Chevron USA, Inc., on soft skills in your
organization – the INFORMS perspective
• InfoNewt: Randy Krum, President, on effective
infographics and data visualizations
• Carrie Beam Consulting: Carrie Beam, Principal, on
soft skills to communicate predictive analytics to the
business community
• Analytics-Based Performance Management:
Gary Cokins, Founder & CEO, on analytics for
decision-making
• CREATE: Stephen Hora, Director, on “two is a party and
three is a crowd; how many experts should be allowed?”
HUMAN RESOURCES ANALYTICS
• Google: Mark Rivera, People Analyst, on HR by the
numbers
• Abbott Analytics: Dean Abbott, President, on hiring and
selecting key personnel using predictive analytics
• IBM Research: Aleksandra Mojsilovic, Manager,
Predictive Modeling & Optimization, on changing the
landscape of workforce management
• PepsiCo: Christopher Shryock, Senior Director,
Workforce Planning & Analytics, on building an HR
analytics function with rigor and pragmatism
INFORMS PRIZES
Franz Edelman Competition & Award
• Boasteel (with Logistics Institute of Northeastern
University; RH Smith School of Business-University of
Maryland; and School of Business and Economics-
Loughborough University, U.K.), “Operations
Research Transforms Boasteel’s Operations”
• Chevron Corporation, “Optimizing Chevron’s
Refneries”
• Dell Inc. (with Dell Global Analytics), Dell’s Retail
Transformation through Analytics”
• Delta Commissioner of Holland (with CPB
Netherlands Bureau for Economic Policy Analysis;
Delft University of Technology; Deltares; HKV
Consultants; Ministry of Infrastructure & the
Environment, The Hague; and Tilburg University),
“Economically Effcient Floor Standard to Protect
the Netherlands against Flooding”
• The Kroger Co. (with Wright State University),
“Simulation and Optimization Improves Pharmacy
Inventory Management at The Kroger Company”
• McKesson Corporation (with IBM T.J. Watson
Research Center), “A Holistic Supply Chain
Sustainability Management Solution”
INFORMS Prize
For effective integration of analytics into
organizational decision making
2012 Winner: Memorial Sloan-Kettering Cancer Center
Daniel H. Wagner Prize
For excellence in operations research practice
2012 Winner: University of Southern California and
United States Coast Guard
Innovative Applications in Analytics Award
For creative developments, applications and
combinations of analytical techniques
Finalists:
• ConEdison, Columbia University and MIT
• General Electric Global Research
• IBM
UPS George D. Smith Prize
For effective and innovative preparation of students in
O.R. practice
Finalists: Lehigh University, MIT and Naval
Postgraduate School
Spreadsheet Guru Competition & Award
Recognizing the best in spreadsheet analysis
WWW. I NF OR MS . OR G 66 | A NA LY T I CS - MAGA Z I NE . OR G
The following question was submitted from a
friend at sea:
“We are having a raffe with numerous prizes
on our ship, which has approximately 2,000
persons on board. The ship is split up into two
groups, “management” (or “khaki”), which con-
sists of about 200 persons, and “non-khaki,”
which make up the remainder of the crew. So
far, only three khakis have been selected as
winners – is that low?”
This is a great question for the Five-Minute An-
alyst because, while it is in the sense of mid-term
exams “poorly posed,” it is a real question, and it is
formulated the way that customers (real people) ask
questions.
As a preliminary step, if we defne a khaki winning
a prize as “success” ( X ), assuming independence
and say that the probability of a khaki winning any
particular drawing is p = 200/2000 = .1 and if we knew
there were N drawings made to the current time, then
we know that the distribution of successes in a fxed
number of independent trials with a constant probabil-
ity of success is binomial. Done, right?
The raffle
BY HARRISON SCHRAMM
This is a great question
for the Five-Minute
Analyst because … it is
a real question, and it is
formulated the way that
customers (real people)
ask questions.
FI VE-MI NUTE ANALYST
MAR CH / A P R I L 2013 | 67 A NA L Y T I C S
The answer is OK except that it
solves the problem he doesn’t have!
Moreover, it falls into a category pro-
fessional analysts should avoid: “true
but not useful.” We could, of course,
go back to our friend and ask, ‘How
many drawings have there been?”
but this requires him to send another
e-mail and that’s difficult for him (for
those with no shipboard experience,
imagine solving this problem for the
Curiosity Mars rover). We’d think we
can give him a “turn-key” answer that
tells him the information he needs and
immediately solves the problem.
Our friend’s real question is: “How
many drawings can we have with only
three khakis being selected before we
determine that the raffe is unfair?” We
are interested in the number of trials re-
sulting in a fxed number of successes;
X is fxed and N - X is a random variable.
This situation is described by the Nega-
tive Binomial Distribution, which counts
the number of failures before a fxed num-
ber of successes. Many of us are already
The Richard E. Rosenthal Early Career
Connection (ECC) is a program of special
events that introduces participants to
well-established researchers and
practitioners for more effective
communication. The program includes
registration to the Conference, as well
as networking events exclusive to ECC
participants.
ECC provides participants with
new perspectives into some
of the most critical problems
facing industry today,
enabling them to broaden
their research agendas. The
goal is for these O.R. leaders
of the future to have an
opportunity, early in
their careers, to
apply their
outstanding
analytical talents
to important
business
problems.
• Nominee's name,
• Email address and telephone number
• Organization and department
• Nominee’s position at the organization
• Nominee's research areas
• Nominee’s year of PhD completion
• A brief paragraph explaining why this
person is being nominated (50-150 words)
NOMINATION DEADLINE:
MARCH 1, 2013
TO NOMINATE A
YOUNG RESEARCHER:
Please send an email to
[email protected],
containing the following
information:
EARLY CAREER CONNECTION EARLY CAREER CONNECTION
ECC is open to researchers from both academia and industry. Participants
must have completed their PhD no earlier than 2007 for this year’s program.
Each participant is nominated by their department chair (for university faculty)
or manager (for industry participants).
Those nominated and selected for this honor will receive an almost 60%
reduction of the conference registration fee to $394.
http://meetings.informs.org/Analytics2013/yrc.html
WWW. I NF OR MS . OR G 68 | A NA LY T I CS - MAGA Z I NE . OR G
familiar with the Geometric Distribution,
which is a special case of the binomial
distribution with X = 1. If you have Excel
2010, you can use NEGBINOM.DIST(),
which gives the option of computing the
cumulative distribution; otherwise, you
can use NEGBINOMDIST(), to get the
PDF and fnd the CDF by keeping a run-
ning tally. So we’re half way to an answer.
The other piece is that we are unsure
what our friend plans to do with the an-
swer; implicitly we need to know some-
thing about our friend’s risk tolerance.
Again, we’re imagining he’s on Mars, so
we’d like to give him an answer in a form
that he can use.
What we send him looks like Figure 1.
Using Figure 1, we
can see that if our friend
wants to be 90 percent
certain that the draw-
ing is unfairly tilted to-
ward non-khakis, there
would have needed to
be more than 48 prizes
awarded to non-khakis
to the three awarded to
khakis. If he wanted to
be 95 percent certain,
the number jumps to 57.
The point of this
month’s column is
twofold: First, to dem-
onstrate that many situ-
ations appear to be unfair when in fact
(at least in statistical terms) they are fair.
Second, to remind analysts in all appli-
cations that the answer not only has to
be “true” but also “useful.” If we can an-
swer our customer’s question, OK, but if
we can answer all questions that might
arise from a given scenario at once (as
we have done here) – even better. As a
side beneft, we have reminded our read-
ers of the power of the negative binomial
distribution, the geometric distribution’s
frequently neglected big brother. ❙
Harrison Schramm ([email protected]
com) is a military instructor in the Operations
Research Department at the Naval Postgraduate
School in Monterey, Calif., and a member of
INFORMS.
FI VE-MI NUTE ANALYST
Figure 1: Cumulative distribution function for the Negative
Binomial Distribution with 3 successes, fxed probability of
success p = .1. This chart allows our friend to decide his risk
tolerance (or required certainty) on the Y-axis, and then look up
how many non-khaki selections would need to be made before we
considered the drawing to be “unfair.”
October 6-9
Minneapolis Convention Center
Minneapolis, Minnesota
Minneapolis – City of Lakes – the
heart of the upper Midwest, looks
forward to hosting you for the
2013 INFORMS Annual Meeting.
General Chair
Robert G. Haight
USDA Forest Service
Program Chair
William L. Cooper
University of Minnesota
Call for Papers
Abstract Deadline: May 15, 2013
Thanks to Our Sponsor!
http://meetings2.informs.org/minneapolis2013
Minn'13 color ad_Layout 1 1/16/13 5:12 PM Page 1
WWW. I NF OR MS . OR G 70 | A NA LY T I CS - MAGA Z I NE . OR G
Constructing a chandelier can
be a tricky undertaking because the
slightest imperfection will unbalance
the chandelier and cause it to be
skewed.
Figure 1 shows a chandelier
constructed from arms, wires and
triangles that hold weights. In order
to perfectly balance the chandelier,
weights must be placed into the tri-
angles. There are nine weights as
follows: 1, 2, 3, 4, 5, 6, 7, 8 and 9kg.
Each triangle can only hold one weight. Assume the
weight of the arms, wires and triangles are negligible.
QUESTION: Where should the weights be placed in
order to perfectly balance the chandelier?
Send your answer to [email protected] by May
15. The winner, chosen randomly from correct an-
swers, will receive an “Analytics - Driving Better Busi-
ness Decisions” T-shirt. Past questions can be found
at puzzlor.com. ❙
John Toczek is the director of Decision Support and Analytics for
ARAMARK Corporation in the Global Risk Management group. He
earned his bachelor’s degree in chemical engineering at Drexel University
(1996) and his master’s degree in operations research from Virginia
Commonwealth University (2005). He is a member of INFORMS.
Chandelier Balancing
BY JOHN TOCZEK
THI NKI NG ANALYTI CALLY
Figure 1: Balancing act: Where should the weights be
placed?
OPTIMIZATION
www.gams.com
Europe
GAMS Software GmbH
[email protected]
USA
GAMS Development
Corporation
[email protected]
http://www.gams.com
GAMS Integrated Developer Environment for editing,
debugging, solving models, and viewing data.
High-Level Modeling
The General Algebraic Modeling System (GAMS)
is a high-level modeling system for mathemati-
cal programming problems. GAMS is tailored for
complex, large-scale modeling applications, and
allows you to build large maintainable models that
can be adapted quickly to new situations. Models
are fully portable from one computer platform to
another.
State-of-the-Art Solvers
GAMS incorporates all major commercial and
academic state-of-the-art solution technologies for
a broad range of problem types.
In a joint effort, the Rail Transport Management
Team and the Scientifc Computing unit at BASF
have developed a model and a solution approach
that provide decision support for the future
structure and size of the rail car feet. The model
consists of a MILP formulation to confgure the feet
structure, and an approximation from inventory
theory to determine the feet size in the presence
of the existing uncertainties. The suggested model
outcome is a considerable reduction in the number
of different rail car types to use and the required
rail car safety stocks, which translates into major
cost savings.
• Data are extracted from Excel-spreadsheets using
GDXXRW and the GAMS GDX-technology.
• Initial solutions for the MILP problems are derived
from a graph coloring problem.
• The solution technique is polylithic and the
mipstart feature of GAMS/CPLEX is used
extensively; 8 GAMS fles act together.
• The GAMS/CPLEX presolve techniques reduce the
mathematical problem size signifcantly.
Transport Logistics at BASF

Sponsor Documents

Or use your account on DocShare.tips

Hide

Forgot your password?

Or register your new account on DocShare.tips

Hide

Lost your password? Please enter your email address. You will receive a link to create a new password.

Back to log-in

Close