ESTIMATION OF SOJOURN TIME DISTRIBUTIONS FOR CYCLIC SEMI-MARKOV PROCESSES IN EQUILIBRIUM
成果类型：
Article
署名作者：
DEWANJI, A; KALBFLEISCH, JD
署名单位：
University of Waterloo
刊物名称：
BIOMETRIKA
ISSN/ISSBN：
0006-3444
DOI：
10.1093/biomet/74.2.281
发表日期：
1987
页码：
281288
关键词：
摘要：
The EM algorithm is used to develop procedures for estimating the interoccurrence distributions when n independent and identically distributed cyclic semi-Markov processes, each being ergodic, irreducible and in equilibrium, are observed over finite windows. This work is an extension of Vardi (1982a, b) who considers a similar problem for renewal processes and develops the RT algorithm for estimation of the common interoccurrence distribution. The results are illustrated using simulations for alternating renewal processes and compared with an extension of an approximate approach of Denby and Vardi (1985).