Professional Documents
Culture Documents
O R G
Data Visualization
Whats next: Seeing is believing
Realization: Big picture of big data
ALSO INSIDE:
2015: Predictions for predictive analytics
Small analytics company takes on Ebola
Case study: Why cloud analytics works
Executive Edge
Bill Gossman, head
of ASA Corp., on
why leadership is
the real gamechanger for using
data analytics
a n a ly t i c s - m a g a z i n e . o r g
IN A FEW
HOURS
MISSION CRITICAL
ENTERPRISE APP
IN A FEW
MONTHS
PUBLISHED
INSTANTLY
TO YOUR ENTERPRISE
OPTIMIZATION
APP STORE
PROOF OF
CONCEPT
IN A FEW
DAYS
OPTIMIZATION APP
IN A FEW
WEEKS
C o n t e n t s
january/february 2015
Brought to you by
Features
44
38
44
52
60
60
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
66
70
Departments
2
8
14
18
22
26
30
34
66
70
74
Inside Story
Analyze This!
Whats Trending
Healthcare Analytics
Viewpoint
Forum
INFORMS Initiatives
Newsmakers
Conference Preview
Five-Minute Analyst
Thinking Analytically
INFORMS Offices
www.informs.org Tel: 1-800-4INFORMS
Executive Director Melissa Moore
Meetings Director Laura Payne
Communications Director Barry List
Headquarters INFORMS (Maryland)
5521 Research Park Drive, Suite 200
Catonsville, MD 21228
Tel.: 443.757.3500
E-mail: informs@informs.org
Analytics Editorial and Advertising
Lionheart Publishing Inc., 506 Roswell Street, Suite 220, Marietta, GA 30060 USA
Tel.: 770.431.0867 Fax: 770.432.6969
President & Advertising Sales John Llewellyn
john.llewellyn@mail.informs.org
Tel.: 770.431.0867, ext. 209
Editor Peter R. Horner
peter.horner@mail.informs.org
Tel.: 770.587.3172
Assistant Editor Donna Brooks
donna.brooks@mail.informs.org
Art Director Jim McDonald
jim.mcdonald@mail.informs.org
Tel.: 770.431.0867, ext. 223
Advertising Sales Sharon Baker
sharon.baker@mail.informs.org
Tel.: 813.852.9942
Analy ze T h i s !
Thoughts on collaboration
and communication
There seems to be
violent agreement
about the value of
collaboration and
communication, and
yet we too often give
it little more than lip
service.
By Vijay Mehrotra
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
Analy ze T h i s !
a n a ly t i c s - m a g a z i n e . o r g
CM
MY
CY
Steve B.
Class of 15
THE COMPETITIVE
ADVANTAGE DEGREE
CMY
a na l y t i c s
j a n u a r y / f e b r u a r y 2 015
11
Analy ze T h i s !
The second of these posts [5], featured a list of 10 data science predictions
for 2015 originally created by the Institute
for Advanced Analytics (http://iianalytics.
com). This list also had a heavy emphasis on collaboration and communication.
12
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
a na l y t i c s
j a n u a r y / f e b r u a r y 2 015
13
Wh at s t r e nd i ng
Top 5 analytics
predictions for 2015
Glenn Wegryn, president of the Analytics Section of INFORMS and principal
at Analytic Impact LLC, offers his top five analytics trends for 2015.
14
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
2015
a na l y t i c s
15
Wh at s t r e nd i ng
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
society no longer a
this area.
increase compliance.
nies will cultivate in-house analytic competencies and many will embrace analytics as
re-invent it.
a na l y t i c s
j a n u a r y / f e b r u a r y 2 015
17
healt h car e a na ly t i c s
By Rajib Ghosh
18
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
manner inside multiple electronic systems that do not talk to each other. Releasing data from such silos is the key to
deriving benefits from the use of healthcare analytics. Until now, providers,
pundits and politicians have blamed electronic health record (EHR) companies for
their product strategy of keeping patient
data locked into their own proprietary
databases. Epic, the leader in market
share among EHR companies, recently
announced that it has hired a lobbyist
firm to fix its image on Capitol Hill. Epic
argues that its product can connect with
CONFERENCE ON
CAREER
CENTER
April 12-14, 2015 | Hyatt Regency Huntington Beach Resort & Spa | Huntington Beach, CA
Find the seasoned professionals you need over 800 analytics professionals expected
Provide your recruitment materials in a casual setting
Arrange discreet on-site meetings in private booths
Enjoy discounted combination pricing with the fall Annual Meeting Job Fair
Enhance your visibility with an ad in Analytics or OR/MS Today
Questions?
a na l y t i c s
19
healt h car e a na ly t i c s
20
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
j a n u a r y / f e b r u a r y 2 015
21
vie w poin t
Cloud analytics
works for me
Using the public
cloud practically
eliminates the time
lapse on infrastructure
procurement.
By Ganesh Moorthy
22
Lets assume that you are asked to analyze a significantly large data set and provide insights. Based
on the data size, you determine that youd require a
clustered and distributed environment to be able to
run the analysis within the pre-decided timeline. You
ask your manager for help and he says to raise a request with IT. You raise the request in the ticketing
system and check with IT. They answer that they dont
have it ready and will need to procure the system first.
You comply and ask how long will the process take?
The IT personnel responds, Four weeks provided
you get the budget approval from your department.
Your timelines have just run amok!
If the above scenario is true in your case, then the
situation is similar to what I have heard from several
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
Set Yourself Apart. Join the Analytics Revolution. Become a CAP Now.
Visit http://www.informs.org/certification or email us at certification@informs.org with any questions.
a na l y t i c s
j a n u a r y / f e b r u a r y 2 015
23
vie w poin t
Based on the infrastructure requirement, we looked into public cloud providers options and decided to go with
Amazon Web Services. It took us just a
couple of days (includes internal approval
process) to get three large capacity nodes
with Linux pre-installed. We wanted a
clean system as we were building a custom natural language processing system
to process the tweets.
It took us one week to deploy Apache
Hadoop, Druid, R statistical tool, Zero
MQ, Python and Java and two weeks to
Custom Analytics on
run through performance benchmark testPublic Cloud
ing and an optimization exercise. In parAs part of an R&D engagement for allel, we also developed and deployed a
one of our clients, we needed to perform custom HTML5 based visualization tool
sentiment analysis on tweets in order for insights.
to determine the problem context, keyAll in all, the entire exercise took a
words, sentiments and topics. The main month and a half from start to finish for
objective of this exercise was to deter- the initial proof of concept. The NLP algomine service-related issues and result- rithm development journey is still undering sentiment of customers. This being a way and being constantly optimized for
social media analytics exercise, the solu- accuracy and performance.
tion was to incorporate aspects of realWhat could have easily been a three
time sentiment complex event processing months exercise if we had waited for
along with data aggregation and historical internal infrastructure procurement, took
descriptive analytics.
only half the time to deliver. We also
We determined that using a Twitter have both data ingestion and insights
streaming service would be the best op- visualization on the public cloud along
tion. This would require constantly listen- with backend processing components
ing to the streaming service and would not to mention, we did not make any
result in collecting and analyzing up to depreciating investments, hence keeping
2,000 tweets a minute.
the bottom line intact.
24
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
Conclusion
Cloud-based analytics or cloud
analytics is an evolution in the making.
We are seeing traditional BI vendors providing cloud-based SAAS-BI services on
one hand, while the others are providing specific niche services in social media analytics using third-party sourced
data. In between, we have big cloud infrastructure providers, providing both infrastructure and analytical platform. On
the most crucial aspects of data security,
most of them have ironclad service-level
ontinuing
ducation
COURSES FOR
ANALYTICS
PROFESSIONALS
FOUNDATIONS OF MODERN
PREDICTIVE ANALYTICS
DATA EXPLORATION
AND VISUALIZATION
a na l y t i c s
j a n u a r y / f e b r u a r y 2 015
25
forum
By Polly MitchellGuthrie
26
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
http://livewebcast.net/INFORMS_AM_Wagner_Prize_2014/pres1/
a na l y t i c s
j a n u a r y / f e b r u a r y 2 015
27
for um
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
Introducing the
H O W M AT U R E I S YO UR O R G ANI Z ATI O N?
https://analyticsmaturity.informs.org
Assess the way your organization uses analytics
Create an improvement plan
Track your progress
Tap the resources of the leading professional
association for advanced analytics professionals
THE INFORMS DIFFERENCE
Reputable
International Leader
Free Access to INFORMS Analytics Maturity Model
Resilient Model That Grows to Meet Your Needs
Begin using the beta version of the INFORMS Analytics Maturity Model.
Just go to https://analyticsmaturity.informs.org. Start today!
For additional information, email informs@informs.org or call 433-757-3500, ext. 560
a na l y t i c s
j a n u a r y / f e b r u a r y 2 015
29
Edelman Award,
continuing education,
IAA finalists
INFORMS announces six finalists for the 2015 Edelman Award
Widely considered
the Super Bowl of
operations research,
as well as INFORMS
highest organizational
honor, the Edelman Award
acknowledges the best
application of high-end
analytics.
30
INFORMS recently announced six finalists representing a diverse collection of applied analytics and
advanced operations research projects that will compete for the 2015 Franz Edelman Award in a series
of judged presentations at the INFORMS Conference
on Business Analytics & Operations Research in Huntington Beach, Calif., in April.
Widely considered the Super Bowl of operations research, as well as INFORMS highest organizational honor, the Edelman acknowledges the best
application of high-end analytics. The winner will
be announced at an evening Edelman Gala held in
conjunction with the conference.
The finalists and their presentations include:
IBM for Predictive Cloud Computing with Big
Data: Professional Golf and Tennis Forecasting
Ingram Micro for End-to-End Business Analytics
and Optimization in Ingram Micros Two-Tier Distribution Business
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
31
32
a n a ly t i c s - m a g a z i n e . o r g
Innovative Applications in
Analytics Award finalists
The Analytics Section of INFORMS
has named three finalists for its Innovative
Applications in Analytics (IAA) Award. The
Sections flagship prize, the IAA recognizes creative and unique developments,
applications or combinations of analytical
techniques used in practice.
The three finalists, chosen by a panel
of judges following a two-month verification process, will present their work at
the INFORMS Conference on Business
Analytics and Operations Research in
Huntington Beach, Calif., in April 2015,
where the winner will be announced.
The three finalists include:
Simulation Approach for Aircraft
Spare Engines & Engine Part Planning,
Jose A. Ramirez-Hernandez, Steven
Oakley, Mei Zhang, Alejandro Scalise,
American Airlines
Intelligent Surgical Scheduling
System, Narges Hosseini, Kalyan
Pasupathy, Yariv Marmor, Thomas
Rohleder, Jeanne Huddleston, Paul
Huddleston, Mayo Clinic Rochester
Fusion Analytics for Public Transport Event Response, IBM Research
and the Land Transport Authority of
Singapore
For more about the competition,
click here.
w w w. i n f o r m s . o r g
6/9/14
9:15 AM
CM
MY
CY
CMY
ne w s m ak e r s
34
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
s Access TO HIGH QUALITY RELEVANT JOB POSTINGS
.O MORE WADING THROUGH POSTINGS THAT ARENT
APPLICABLE TO YOUR EXPERTISE
s 0ERSONALIZED JOB ALERTS NOTIFY YOU OF RELEVANT
JOB OPPORTUNITIES
s #AREER MANAGEMENT n YOU HAVE COMPLETE
CONTROL OVER YOUR PASSIVE OR ACTIVE JOB SEARCH
5PLOAD MULTIPLE RESUMES AND COVER LETTERS
ADD NOTES ON EMPLOYERS AND COMMUNICATE
ANONYMOUSLY WITH EMPLOYERS
s !NONYMOUS RESUME BANK PROTECTS YOUR CONlDENTIAL
INFORMATION 9OUR RESUME WILL BE DISPLAYED FOR
EMPLOYERS TO VIEW %8#%04 YOUR IDENTITY AND
CONTACT INFORMATION WHICH WILL REMAIN CONlDENTIAL
UNTIL YOU ARE READY TO REVEAL IT
s 6ALUE
ADDED BENElTS OF CAREER COACHING RESUME
SERVICES EDUCATIONTRAINING ARTICLES AND ADVICE
RESUME CRITIQUE RESUME WRITING AND CAREER
ASSESSMENT TEST SERVICES
http://careercenter.informs.org
na l y t i c s
a dave
r t i s e r INDEX
PAGE #
ADVERTISER
POWERED BY
j a n u a r y / f e b r u a r y 2 015
35
ne w s m ak e r s
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
j a n u a r y / f e b r u a r y 2 015
37
EXECUTIVE E D GE
38
a n a ly t i c s - m a g a z i n e . o r g
39
EXECUTIVE E D GE
that is not so readily apparent. The biggest barrier is the assumption that analytics is a data science and technology problem when its really a
leadership and managerial problem, says Florian
Zettelmeyer, director of the data analytics program
at the Kellogg School of Management, Northwestern University. The managerial class has to understand the problems that need to be solved and
how analytics can help solve them.
When leadership fails to implement a decision
management strategy and convey its importance
throughout the company, it unwittingly bypasses
a major growth opportunity. Data constantly flows
into and around an organization. Leveraging its
specific components requires analytical execution
and deployment in areas of maximum benefit by
a committed staff. This is the essence of opportunity management. Effective decision management uses analytics to achieve better outcomes
from customer interactions. The value, of course,
depends on how well all these factors are applied.
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
the size of the company. Larger corporations generally engage statisticians in addition to a business intelligence team as
part of their overall strategy. Most SMBs
cannot afford to contract with statisticians,
but have other avenues to maximize business intelligence, most notably automated
programs. Some add BI expertise to their
staffs, but are uncertain what their next big
step should be. At this stage, leadership
has to be prepared to resolve inevitable
internal conflicts about analytical direction.
What is needed is a strategy to integrate analytics into decision management beginning with two very simple and
achievable practices: focus and common
sense. You must have a long-term plan
and a strategy that works with it, says
David Hufnagel, chief operations officer
and vice president of Congressional Federal Credit Union (CFCU). If you dont
work in concert with the plan, it becomes
inoperable with no results. The credit
union has relied on automated analytics
for the last three years to drive a proactive outreach to its 45,000 members. Its
not about being unique with each individual, Hufnagel says. Its about identifying
what is relevant to our members.
The CFCU experience is an example
of focus and common sense for maximizing the value of data analytics through
technological innovation. The idea is to
have the technology focus on data that
a na l y t i c s
41
EXECUTIVE E D GE
42
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
Huntington Beach
APRIL 12-14, 2015
Present your work at the conference known for real-world analytics and save on registration.
Case studies, best practice examples, and academic research with a practitioner orientation
are all welcome.
POSTER PRESENTATION SUBMISSIONS DUE
January 16, 2015
http://meetings.informs.org/analytics2015
data v is ua li z at i o n
a n a ly t i c s - m a g a z i n e . o r g
a na l y t i c s
45
data v is ua li z at i o n
Figure 2:
Pre-Internet
sociogram.
Source:
Journal
of Social
Structure
Figure 3:
Internet-age
sociogram.
Source: Leadership
Learning Community
46
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
47
data v is ua li z at i o n
a n a ly t i c s - m a g a z i n e . o r g
The vast majority of data visualizations today are two-dimensional. However, thats changing with creative use of
color and size, combination of space and
time, and advanced computer graphics.
For instance, neuroscientists Emmanuelle
w w w. i n f o r m s . o r g
Tens of billions of devices will be connected to the Internet in the next decade.
From smart appliances and wearables to automobile sensors and environmental
monitors, the Internet of Things will provide unprecedented insight into whats
happening around us. High-throughput, interconnected data streams will help us
improve safety, drive operational efficiencies and better understand consumer
demand.
In the words of Kevin Ashton, who first coined the term the Internet of Things
in his seminal 2009 RFID Journal article, The Internet of Things has the potential
to change the world, just as the Internet did. Maybe even more so.
Network Theory
Network Theory builds on Graph Theory, which applies algorithms to understand and model pair-wise relationships between objects. Network Theory
examines relationship symmetry, with the existence of asymmetric relationships providing grounds to predict the likely spread of information (social network analysis), dissect complex disorders (biological network analysis), find
the shortest path between two points (network optimization) and identify target
objects based on their behavior (link analysis).
Complexity Theory
a na l y t i c s
j a n u a r y / f e b r u a r y 2 015
49
data v is ua li z at i o n
Multidimensional Visualization
The adage a picture is worth a thousand words gained credence from our
ability to process visuals more easily than text. Visualization has also been shown
to improve learning and recall, and can portray complex concepts and relationships more easily than can text. Recent developments in computer graphics are
making possible visualizations that enable the integration, manipulation and exploration of dynamic multidimensional data sets. Multidimensional visualizations
allow users to not only examine data from new perspectives but also interact with
it more effectively.
50
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
As the world becomes increasingly interconnected and interdependent, opportunities to generate value through data
visualization will only increase. The Internet of Things will have a profound effect
on the role that data visualization can play
in organizations and society, improving
our ability to understand how humans and
machines interact with each other and the
environment. Application of evolving cognitive frameworks, such as Network and
a na l y t i c s
j a n u a r y / f e b r u a r y 2 015
51
data v is ua li z at i o n
Data visualization:
The big picture
of big data
By Nana S. Banerjee
a n a ly t i c s - m a g a z i n e . o r g
Source: The User Illusion: Cutting Consciousness Down to Size, by Tor Nrretranders (Penguin Press Science)
brain and how understanding those processes can provide profound insights.
In his best-selling 2012 book Thinking,
Fast and Slow, Nobel laureate Daniel
Kahneman introduces the terms System
1 and System 2. The terms differentiate
between the information processing that
occurs in the human subconscious and
conscious minds. System 1 addresses
the functions that are uncontrolled and
effortless. System 2 comprises functions that are controlled and require
effort to engage. In action, System 1
allows us to instantaneously recognize
facial expressions visual processing.
In contrast, System 2 allows us to make
complex decisions or apply reason.
A little more than a decade before
the release of Kahnemans book, Danish
a na l y t i c s
53
data v is ua li z at i o n
Realizing how quickly we understand and internalize what we see is at the foundation of what
makes data visualization such an important aspect of how we analyze information and make
better decisions. That said, the mechanisms behind data visualization create a powerful tool to
design effective visualizations to suit any context
whether that tool is a simple, static bar chart or
something vastly more complex, multidimensional
and interactive. As such, the science behind data
visualization ranges from the fundamentals of how
we literally see to the complexities of cognitive
psychology.
Combining the science with the art how best
to portray the intent of any particular visualization
winds up somewhere on a curve between presentation and exploration. The difference between
presentation and exploration can be described as
the difference between presenting a known story
in a data set using analysis and exploring a notyet-understood data set using a visual examination. Henry David Thoreau said, Its not what you
look at that matters, its what you see. With data
visualization, the significance of the quote is quite
literal. Its a fully formed discipline that requires
multiple skills among them, the knowledge of
statistics, ideas of space, design and topography,
and a deep subject matter expertise in the sector
being served.
From Theory to Practice
Currently, for any company that deals with a
titanic amount of data, data visualization is and
will remain an absolutely fundamental tool. Verisk
54
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
55
data v is ua li z at i o n
Figure 2: A detailed inundation map of the New England coastline depicted in Figure
2 shows the surge footprint of Hurricane Sandy in blue. The orange circles represent
clustered locations by actual number, with the largest circles containing the most individual locations. Where the blue and orange overlap, the map illustrates where Sandy
had the greatest impact. Such maps help companies determine the extent of floods
and resulting losses.
Source: AIR Touchstone zoomed-in surge
a n a ly t i c s - m a g a z i n e . o r g
57
data v is ua li z at i o n
58
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
JOIN
Analy tic s i n Ac t i o n
a n a ly t i c s - m a g a z i n e . o r g
(ICG), had a conversation with company CEO David Waldrop about the Ebola
epidemic. They each expressed concern
about the scale of the human suffering,
but also about facets of security and public health that could limit the spread of
the epidemic. Given the nature of their
earlier work in real-time streaming data
analytics, they eventually began speaking
about various ways one might apply data
w w w. i n f o r m s . o r g
61
trackin g e bo la
may be boarded, inspected and possibly quarantined. This scenario created compelling incentives for non-compliance with disease reporting,
as ship owners face major commercial revenue
losses if their ships are quarantined, or if contract
conditions are not met owing to delays, or if the
ship is diverted and fails to reach the port specified in a bill of lading. Taking this situation into
account, Waldrops experiment with LUX helped
address some issues of public safety in ports.
Two federal agencies looked at the information. It brought the maritime aspect of the
problem to their attention and demonstrated a
relatively easy way to pay attention to it. They
asked about the data sources and began monitoring them. The ships were monitored on the
LUX watch dashboard. This procedure provided
several days of advance warning about any ship
that had visited West Africa and was headed to
a U.S. port.
Ebola Maritime Vector Threat
To mitigate the Ebola threat, LUX uses two
real-time data streams to perform analysis. The
first is data streamed from internationally mandated [1] automatic identification system (AIS)
reporting. AIS, installed on all merchant ships
in the world, is especially relevant to monitoring
ships traveling to and from West Africa ports. The
second source is data generated by the Global
Data on Events, Location and Tone (GDELT)
project [2].
The LUX Ebola Maritime Vector monitoring
methodology works as follows:
62
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
Figure 1: Illustration of the LUX user interface. The outer AOI provides alerts on
arrivals and departures from the zone of concern. The inner AOI monitors activity in
and around port areas.
a na l y t i c s
63
trackin g e bo la
a n a ly t i c s - m a g a z i n e . o r g
Subscribe to Analytics
Its fast, its easy and its FREE!
Just visit: http://analytics.informs.org/
w w w. i n f o r m s . o r g
ting
n
e
v
n
i
e
R alue
V
HEALTHCARE 2015
NASHVILLE TENNESSEE
Abstract Submissions Now Open!
http://meetings.informs.org/healthcare2015/abstract-submission
INFORMS Healthcare 2015 invites submissions on research and
applications in the healthcare arena. We welcome submissions in
either oral or poster formats, which will be reviewed by the
Committee. There is a limit of one oral presentation per
presenting/lead author. Authors may present a poster as
well as one oral presentation.
JULY 29-31
OMNI HOTEL
GENERAL CHAIRS
Robert Dittus, Vanderbilt University
M. Eric Johnson, Vanderbilt University
PROGRAM CHAIR
Vikram Tiwari, Vanderbilt University
INVITED SESSIONS CHAIR
Eva Lee, Georgia Institute of Technology
http://meetings.informs.org/healthcare2015
ADVISORY COMMITTEE
Dr. Warren Sandberg, Vanderbilt University
R. Lawrence Van Horn, Vanderbilt University
co n fer e n c e p r e v i e w
By Manoj Chari
66
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
67
co n fer e n c e p r e v i e w
68
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
TM
b3analytics.com/data
The Data & Statistics Store is a onestop location for marketing and
business analysts, data scientists, and analytics professionals to gain
access to preintegrated demographic, socioeconomic, and behavioral data
Five - M in u t e A na lyst
Battleship
The research
question is: How
can we tell if
another player is
being dishonest?
By Harrison
Schramm, CAP
70
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Figure 1: A scheme for covering the Battleship board in 33 shots. There are similar
schemes for covering the board in 50 and 20 shots.
5.3. This is a minor but noticeable difference in the number of times to first hit. I
would begin to seriously suspect (95 percent confidence) that my opponent was
cheating if I did not score a hit in the first
15 attempts, and would be almost certain
(99 percent confidence) he was cheating
if I did not score a hit in the first 22 shots.
Can we do better? Yes! Each ship is
rectangular, and can only inhabit squares
that are along ranks or files (like rooks
in chess). The way to handle this is to
pick shots along diagonals (like bishops
j a n u a r y / f e b r u a r y 2 015
71
Five- M in u t e A n a lyst
possibility of missing the two-square corvette ship, but will certainly hit all the others. This is an area ripe for further analysis.
Defending along the rows: What does
it mean to be maximally random? There
are different interpretations, but one way
might be to make the groupings along
the rows and columns of the board fit a
Poisson distribution with =1.7. A perfect
fit is not possible because at least one of
the rows or columns will have a 5 due
to the carrier, which only has a .3 probThe formulae for the other spacing are
ability of occurrence. This turns out to be
similar.
an optimization problem that exceeds the
We can compare these approaches,
bounds of the Five-Minute Analyst. This
and see that they all perform remarkably
problem is very similar to solving Sudoku
similar; the five-space approach is slightly
with optimization. Below is a candidate
better for achieving the first hit.
solution and its performance measured
by the Poisson metric.
In conclusion, this piece has brought
up more areas for analysis on this topic
than presented answers. Hopefully you
will think differently about Battleship next
time you play.
Bonus: If you dont have your computer handy while you are figuring out
Figure 3: Performance of various skipwhich strategy has the greatest rate of
ping schemes for scoring the first hit in
improvement, you can appeal to calculus.
Battleship.
The proposed strategies are of the form
We have only considered the time until
the first hit. While beyond the bounds of
. Differentiate, yielding
this analysis, my sense is that skip two,
as depicted in Figure 1, is the optimum
and evaluate at a convenient
game playing strategy. This strategy has a place, say, n = 1 .
72
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
1
1
1
1
1
1
1
1
1
1
a na l y t i c s
j a n u a r y / f e b r u a r y 2 015
73
Thin k in g A na ly t i ca lly
Electrifying
By John Toczek
John Toczek is the senior director
of Decision Support and Analytics for
ARAMARK Corporation in the Global
Operational Excellence group. He
earned a bachelor of science degree
in chemical engineering at Drexel
University (1996) and a masters
degree in operations research from
Virginia Commonwealth University
(2005). He is a member of INFORMS.
74
a n a ly t i c s - m a g a z i n e . o r g
w w w. i n f o r m s . o r g
OPTIMIZATION
GENERAL ALGEBRAIC MODELING SYSTEM
High-Level Modeling
The General Algebraic Modeling System (GAMS) is a high-level
modeling system for mathematical programming problems.
GAMS is tailored for complex, large-scale modeling applications,
and allows you to build large maintainable models that can be
adapted quickly to new situations. Models are fully portable from
one computer platform to another.
State-of-the-Art Solvers
GAMS incorporates all major commercial and academic state-of-
sales@gams.com
www.gams.com