A Tutorial on
“New” Analysis Techniques:
Understanding & Applications
25-27
January 2005
Johns Hopkins University/Applied Physics Lab
Laurel, Maryland
Terms
of Reference
1. Purpose
The
purpose of this tutorial mini-symposium is to further the professional
development of MORS’ members by offering a series of educational sessions on
“new” scientific topics that are relevant to the profession of military
operations research. This mini-symposium will focus on a variety of techniques,
some of which may be truly new, and/or some older, more established techniques
that are being applied in “new” ways. Tutorial topics will include: Network
Interdiction Modeling with Military Applications; Multi Resolution Modeling; and
Advancements in Large Scale Experimental Design.
2. Background
This
mini-symposium will be the third in a series of MORS-sponsored tutorials and
follows the highly successful tutorial on New Techniques: A Better Understanding of their Application to Analysis,
held in the fall of 2002, and the earlier workshop on Warfare Analysis and Complexity, conducted in September 1997. Both
of these meetings were extremely well attended, received laudatory comments from
the participants and MORS’ Sponsors, and subsequently enabled analysts to use
the tools and techniques to better inform decision-makers.
MORS
has a legacy of conducting education and professional development Mini-Symposia
for the education of its members. The complexity and scope of challenges that
are confronting today’s decision-makers have served to broaden the set of
analytical tools and techniques that should be considered by the practitioners
of military operations research. As such, it is appropriate that MORS convene
this mini-symposium to ensure that MORS’ members have the opportunity to
maintain their currency. The target audience for this mini-symposium includes
both OR analysts and OR managers.
3. Type
of Special Meeting
This
will be a three-day Mini-Symposium. Each day will be designed to focus on a
distinct Tutorial topic. The Tutorials will be on specific analytical
techniques—e.g., Network Interdiction Modeling with Military Applications;
Multi Resolution Modeling; Advancements in Large Scale Experimental Design —
and not on particular tools (e.g., Model-X, Model-Y).
The
format of each Tutorial will be theoretical - plenary presentations in the JHU/APL
Kossiakoff Center auditorium interspersed with hands-on exercises in the
adjacent classrooms. It is intended that attendees would be provided with the
opportunity to not only grasp the theoretical foundation of the “new”
technique but also have the opportunity to apply it to a real world, military OR
problem set. It is also intended that attendees be provided with a set of
training materials from the Special Meeting such as a CD that might contain at a
minimum the theoretical foundation of the techniques and a set of application
problems.
Note
that classrooms in the Kossiakoff Center will have to be vacated by 1530 due to
the conduct of Johns Hopkins University evening classes. Mini-symposium
participants will return to the auditorium for a final session on each day.
4.
Dates
25-27 January 2005
5. Location
Kossiakoff Center, Johns Hopkins University Applied Physics Laboratory,
Laurel, MD
6. Classification
Unclassified
7. Organizing
Committee
Co-Chairs:
Mr. Ted Smyth, PP and Mr. Bill Kroshl (JHU/APL)
Sponsor
Interest: Dr. Jackie
Henningsen, FS, AFSAA; Mr. Walt Hollis, FS, DUSA(OR);
Mr. Eric Coulter, OSD; and, the Joint Staff (represented by the J8)
Directors from Academia:
Dr. Andy Loerch (GMU) MORS President; LTC Willie McFadden, PP, LTC Mike Kwinn, and Dr. Niki Goerger
(USMA); Dr. Steve Pilnick (AD-NPS)
Other
Directors: Mr. Pat McKenna (STRATCOM); Mr Greg Keethler (USAF SPACECOM);
Mr. Robert Holcomb (IDA); Mr. Kirk Michealson (Lockheed Martin); Mr. Terry
McKearney (The Ranger Group)
Prospective Tutorial Leads:
Tutorial #1
– Network Interdiction Modeling
with Military Applications (Naval Postgraduate School faculty members)
This
tutorial will discuss the theory and practice of modeling the attack and defense
of various types of networks. Highlights of the session include:
converting publicly available but unorganized data into useable data structures
that allow incisive analysis; an overview of network models, (shortest path, max
flow, min-cost flow, multicommodity flow, and spanning trees), guided by
applications to minimum-risk flight paths, fuel distribution, and logistics
systems; in-depth case studies of two problems with underlying network models
– delaying a covert nuclear weapons development program, and prepositioning
missile defense platforms to provide optimal protection against theater
ballistic missiles; and finally, the use of Excel and Java to solve these
problems. In modern warfare, the advantage will go to the force that
understands both their own networks and those of the adversary, and has the
tools to use that information to effectively attack the enemy and defend
oneself.
Tutorial #2
– Multi Resolution Modeling (Dr.
Paul Davis, RAND)
This
tutorial will discuss the theory and practice of multiresolution modeling (MRM),
to include work with families of models and war games, and the exploratory
analysis that is enabled by having good low-resolution models as part of
one's kit bag. Such exploratory
analysis needs to be a key element of capabilities based planning. The tutorial will cover theory and applications of MRM and
exploratory analysis, to include applications linking low-resolution analysis to
work at the individual-tank level in air-land combat or to work at the system
level in missile defense problems.
Tutorial #3
– Advancements in Large Scale
Experimental Design (TRAC-Monterey personnel/Naval Postgraduate School faculty members)
The
United States Department of Defense uses simulation models to support its
decision-making process. Defense
analysts need experimental designs capable of efficiently searching an intricate
simulation model that has a high-dimensional input space characterized by a
complex response surface (substantial non-linearities may be prevalent).
To efficiently explore these simulations, the experimental designs should
have the
following desirable characteristics:
