报告题目:
Bayesian empirical likelihood for accelerated failure time models with right censored
报告摘要:
By introducing a prior, we construct the Bayesian empirical likelihood function based on the Buckley-James estimator. By using the MH algorithm, we conduct some simulation studies to compare the performance of the Bayesian empirical likelihood and the Bayesian estimator and the B-J estimator and the empirical likelihood. The simulation results demonstrate that Bayesian empirical likelihood performs better than the Bayesian estimator and the B-J estimator and the empirical likelihood practically. Finally, we apply this proposed method to analyze a example, then obtaining a narrower confidence interval of model parameters.
报告时间:2021年7月8日上午9点
报告形式:腾讯会议
会议链接:https://meeting.tencent.com/s/uxpT5ENGllCU
会议 ID:442 815 736
主办单位:数学与统计学院