Files and links (1)
Free to read and download, Open Access
Dissertation
TIGER: an unsupervised machine learning tactical inference generator
University of Iowa
Doctor of Philosophy (PhD), University of Iowa
Summer 2009
DOI: 10.17077/etd.eo2ka3hi
Abstract
<p>We present here TIGER, a Tactical Inference Generator computer program that was designed as a test-bed program for our research, and the results of a series of surveys of Subject Matter Experts (SMEs) testing the following hypotheses:</p>
<p><em>Hypothesis 1: There is agreement among military experts that tactical situations exhibit certain features (or attributes) and that these features can be used by SMEs to group tactical situations by similarity.</p>
<p>Hypothesis 2: The best match (by TIGER of a new scenario to a scenario from its historical database) predicts what the experts would choose. </em></p>
<p>We have conducted three surveys of SMEs and have concluded that there is, indeed, a statistically significant confirmation of Hypothesis 1, that there is agreement among military SMEs that tactical situations exhibit certain features (or attributes) and, that these features can be used to group, or identify, similar tactical situations. The statistical confidence level for this confirmation of Hypothesis 1 is greater than twice the prior probability.</p>
<p>In order to test Hypothesis 2 we constructed, after SME survey analysis, a series of algorithms, which we present here, for the analysis of SME identified tactical features (or attributes) including: <em>interior lines, restricted avenues of approach, restricted avenues of attack, slope of attack, weighted force relationships and anchored or unanchored flanks</em>. Furthermore, the construction, and implementation, of these algorithms, required the design and implementation of certain `building block' algorithms including: <em>range of influence, optimal FindPath, ComputeGroupsByThreshold </em>and <em>ComputeGroupsByNumber. </em></p>
<p>We further present an overview of TIGER, itself, and the built-in utilities necessary for creating three-dimensional tactical situations, complete with terrain, elevation and unit types as well as our implementation of Gennari, Fisher and Langley's CLASSIT classification system.</p>
<p>Lastly, we present TIGER's classification of twenty historical tactical situations and five hypothetical tactical situations and the SME survey results of TIGER's classification that resulted in TIGER correctly predicting what the SMEs would choose in four out of five tests (using a one sided Wald test resulted in p = 0.0001 which is statistically significant).</p>
Details
- Title: Subtitle
- TIGER: an unsupervised machine learning tactical inference generator
- Creators
- David Ezra Sidran - University of Iowa
- Contributors
- Alberto Maria Segre (Advisor)Steven Bruell (Committee Member)Ted Herman (Committee Member)Juan Pablo Hourcade (Committee Member)Joseph Kearney (Committee Member)Rosemary Moore (Committee Member)
- Resource Type
- Dissertation
- Degree Awarded
- Doctor of Philosophy (PhD), University of Iowa
- Degree in
- Computer Science
- Date degree season
- Summer 2009
- Publisher
- University of Iowa
- DOI
- 10.17077/etd.eo2ka3hi
- Number of pages
- xi, 151 pages
- Copyright
- Copyright © 2009 David Ezra Sidran
- Comment
This thesis has been optimized for improved web viewing. If you require the original version, contact the University Archives at the University of Iowa: https://www.lib.uiowa.edu/sc/contact/.
- Language
- English
- Description bibliographic
- Includes bibliographical references (pages 145-151).
- Academic Unit
- Computer Science
- Record Identifier
- 9983777242802771
Metrics
388 File views/ downloads
1026 Record Views