- Thursday, January 1, 2009
- Published at:Applied Mathematical Sciences, Vol. 3, 2009, no. 1, 27 - 42
In this paper, we consider two types of failure data namely grouped and non-grouped, which can be motivated from a linear mixed degradation model. We propose a Bayesian approach to estimate the parameters of the time-to-failure distribution and its percentiles. A simulation study conducted to study the performance of the proposed method showed that in terms of the mean squares error and the length of the bootstrap confidence intervals, the behavior of the proposed method is satisfactory. Also, it showed that the Bayesian approach with non-grouped data is better than the Bayesian approach with a grouped data. Application to a real data set is given.