Non-Parametric Estimation of Component Life Time Distributions of a Parallel System

Abstract

In many practical situations dealing with parallel systems, the failure times of the components are not observable unless that is the last one resulting in the failure of the system. Based on observation on the life time of the system, nonparametric estimation of the life time distributions of the components is considered. Dealing with a parallel system with two components, a competing risks framework is developed and an algorithm of the EM type for maximum likelihood estimation is obtained. The method is illustrated with a simulated data set.

Parametric Estimation of Quality Adjusted Lifetime (QAL) Distribution in Simple Illness-Death Model

“…Starting from the healthy state, an individual may enter an intermediate state before ending in death, or have transition directly to death. Applications can be found in animal carcinogenicity studies (Kodell and Nelson, 1980), medical studies involving human subjects, for example in HIV/AIDS research (Kalbfleisch and Lawless, 1989), and in industrial applications with machine faults (Dewanji and Dhar, 1993).…”