Mixed-Effects Logistic Regression Model for Missile Reliability Prediction

Award Information
Agency: Department of Defense
Branch: Army
Contract: W31P4Q-04-C-R113
Agency Tracking Number: A032-2578
Amount: $69,976.00
Phase: Phase I
Program: SBIR
Awards Year: 2004
Solicitation Year: 2003
Solicitation Topic Code: A03-152
Solicitation Number: 2003.2
Small Business Information
500 West Cummings Park - Ste 3000, Woburn, MA, 01801
DUNS: 859244204
HUBZone Owned: N
Woman Owned: N
Socially and Economically Disadvantaged: Y
Principal Investigator
 Ssu-Hsin Yu
 Group Leader
 (781) 933-5355
Business Contact
 Raman Mehra
Title: President
Phone: (781) 933-5355
Email: rkm@ssci.com
Research Institution
In this Phase I SBIR project we propose to use a Linear Mixed Effects Logistic Regression Model for missile reliability assessment. An effective model candidate for missile reliability assessment must be flexible enough to incorporate the different nature of covariates, yet simple enough to allow practical methods for parameter estimation using a realistic amount of sample data. A Logistic Regression Model is particularly suitable for modeling dichotomous data such as pass/fail in missile tests. The main advantage of a Mixed-Effects Model is that it provides a systematic method of aggregating a large numbers of effects of different nature into a manageable number of random effect terms in the model. This in turn allows us to estimate the model parameters efficiently. By combining the desirable features of these two models, we expect the proposed model will perform favorably compared to standard linear regression models. During this Phase I project we will obtain test data, determine the appropriate model structure and estimate the model parameters, and evaluate the model prediction performance to demonstrate the feasibility of the proposed model. We will also investigate the use of the Expectation-Maximization algorithm to cope with the issue of incomplete covariates data. Dr. Alyson Wilson and Dr. Nicholas Hengartner from Los Alamos Lab will provide consulting support on the choice of model structure and covariates, and parameter estimation issues during the Phase I period.

* Information listed above is at the time of submission. *

Agency Micro-sites

SBA logo
Department of Agriculture logo
Department of Commerce logo
Department of Defense logo
Department of Education logo
Department of Energy logo
Department of Health and Human Services logo
Department of Homeland Security logo
Department of Transportation logo
Environmental Protection Agency logo
National Aeronautics and Space Administration logo
National Science Foundation logo
US Flag An Official Website of the United States Government