|
Activity Number:
|
228
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Defense and National Security
|
| Abstract - #305436 |
|
Title:
|
An Efficient Approach for Finding Optimal Resource Allocations in a Missile Reliability Study
|
|
Author(s):
|
Jessica Chapman*+ and Max Morris and Christine Anderson-Cook
|
|
Companies:
|
St. Lawrence University and Iowa State University and Los Alamos National Laboratory
|
|
Address:
|
127 Bewkes Hall, Canton, NY, 13617,
|
|
Keywords:
|
Resource allocation ; MCMC ; System reliability
|
|
Abstract:
|
Data collection planning is an important step in the experimental design process. We address data collection planning in the context of collecting second-stage data in system reliability studies (often referred to as resource allocation). A Bayesian hierarchical model is used to make initial system reliability assessments. Due to the complexity of the model, Markov Chain Monte Carlo methods are necessary. Previous approaches for finding an optimal resource allocation have been computationally intensive, requiring repeated analyses via MCMC. We have introduced a computationally efficient approach for evaluating candidate resource allocations that requires only a single MCMC analysis. We demonstrate our approach by employing genetic algorithms to search for an optimal resource allocation for collecting additional data to assess the reliability of an air-to-air heat-seeking missile.
|
