Semiparametric Bayesian analysis of high-dimensional censored outcome data  

作　　者：Chetkar Jha Yi Li Subharup Guha 

机构地区：[1]Department of Statistics,University of Missouri,Columbia,MO,USA [2]Department of Biostatistics,University of Michigan,Ann Arbor,MI,USA

出　　处：《Statistical Theory and Related Fields》2017年第2期194-204,共11页统计理论及其应用（英文）

基　　金：National Science Foundation[grant number DMS-1461948].

摘　　要：The Surveillance, Epidemiology and End Results (SEER) cancer database contains survival data forUS individuals diagnosed with cancer. Semiparametric Bayesian methods are computationallyexpensive to fit for such large data-sets. This paper develops a cost-effective Markov chain MonteCarlo strategy for censored outcomes to fit a semiparametric bayesian analysis of SEER data ofNew Mexico. We use an accelerated failure time model, with Dirichlet process random effectsfor inter-subject variation, and intrinsic conditionally autoregressive random effects for spatialcorrelations. The results offer insights into differences in breast cancer mortality rates betweenethnic groups, tumor grade and spatial effect of counties.

关 键 词：ICAR models data squashing Dirichlet process generalised Pólya urn process big data 

分 类 号：R73[医药卫生—肿瘤]