Speaker: Gideon Zamba (U. Iowa)
Title: A Semi-parametric Random-cell type of goodness-of-fit Test when Observations are Recurrent
Abstract: Goodness-of-fit of the distribution function governing time to occurrence of recurrent events is considered. We develop a chi-square type of test based on the nonparametric maximum likelihood estimator (NPMLE) of the inter-event time distribution. The test is based on the minimum chi-squared estimator of a parametric family and compares the estimated parametric null to the NPMLE on k partitions of a calendar time over a study monitoring period. Small sample and asymptotic properties of the proposed statistic are investigated. Simulation results for Weibull lifetime models are discussed and large sample properties of the test statistic are established using empirical process tools. The approach is then applied to jet planes air conditioning system failures.