期刊名称：Brazilian Journal of Probability and Statistics
印刷版ISSN：0103-0752
出版年度：2008
卷号：22
期号：1
页码：23-33
出版社：Brazilian Statistical Association
摘要：In this paper we discuss a general parametric class of models for analyzingmultiple failure data where the study subjects can experience repeated events. The ideais to model the global time, combining the two time scales, total time and interval time(intervals between successive events), and the event counts in a hybrid model, and todecide about their appropriateness in the light of the data. The general framework allowsa broad class of hazard or intensity models including the Poisson and renewal processmodels as very special cases. Simulation results suggest that the test procedures performwell even for small and moderate sample sizes. A Real dataset illustrate the methodology
