Date of Award

2017

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Computer Science : Modeling and Simulation

Committee Chair

Mikel D. Petty

Committee Member

Harry Delugach

Committee Member

David Moody

Subject(s)

Bayesian statistical decision theory, Computer simulation, Sequential analysis

Abstract

Testing for large and complex defense systems can be extremely expensive. A testing strategy is developed to leverage Bayesian Experimental Design concepts within the Model-Test-Model paradigm. Two utility functions are implemented to search the input space and select the test which maximizes the expected information gain per test. A priority-based strategy focuses on individual performance measures in order of importance. A weighted-sum strategy maximizes the information gained across all performance measures, with emphasis placed on higher priority measures which are further from their accuracy specifications. Monte Carlo simulations show that both strategies provide a statistically significant improvement over the baseline random selection strategy. The weighted-sum strategy is found to be more computationally intensive but is still recommended over the alternative due to its better performance in certain cases.

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.