Statistical Inference for Masked Data

This article considers the situation in which a system consists of k components in series and a defect in any component causes a system malfunction. When a system malfunction occurs, test procedures restrict the cause to some subset of the k components. When the subset consists of more than one component, this phenomenon is termed masking. The presentation will first focus on the case where defects have a successs/fail structure. An industrial example will be used to motivate our interest in this problem. Typically masking introduces two types of problems. First, it is desirable to estimate the diagnostic probability, i.e. the probability, given a specified malfunctioning subset, that each of the masked components is the defective one. Second, one would like to estimate the individual component defect probabilities. Our article discusses these problems and derives two stage procedures for estimation and inference. Further, we extend these results to the case where system failure times are observed and show how under a proportional hazard assumption, nonparametric estimation of the component failure distributions and the diagnostic probabilities can be carried out.

By: Benjamin Reiser (Univ. of Haifa, Israel), Emmanuel Yashchin and Betty J. Flehinger

Published in: Nonlinear Analysis - Theory Methods and Applications, volume 30, (no 7), pages 4425-32 in 1997

Please obtain a copy of this paper from your local library. IBM cannot distribute this paper externally.

Questions about this service can be mailed to reports@us.ibm.com .