Your browser does not support JavaScript!
2016-05-27(五),主講人:王維菁教授 (國立交通大學 統計所)



統 計 學 研 究 所


專 題 演 講


講 題: Nonparametric Inference for Recurrent Events Data under Competing Risks 
演講者: 王維菁教授 (國立交通大學 統計所)
時 間: 105年05月27日(星期五)10:40 - 12:10  (上午10:20 - 10:40茶會於統計所821室舉行)
地 點: 綜合三館837室
摘 要: The work is motivated by a medical study on peritonitis in Taiwan, where patients on peritoneal dialysis might experience two types of peritonitis (Gram-positive or non-Gram-positive), each of which could recur. Furthermore a terminal event happened if a patient was transferred to permanent hemodialysis or died. Here we are interested in studying the cause-specific gap times between successive events. Treating the two types of peritonitis and the terminal event as competing risks, we consider nonparametric inference of the marginal cumulative incidence function and cause-specific hazard function, both of which are nonparametrically identifiable. The technique of inverse-probability-of-censoring weighting is employed to handle the problem of induced dependent censoring, a phenomenon that arises in analysis of the second and subsequent gap times. Large-sample properties are well studied and then utilized to develop several useful testing procedures. The proposed methods are examined via simulations and demonstrated by analyzing the peritonitis data mentioned above.

敬請公佈                                 歡迎參加

  • 友善列印
  • 分享
Voice Play