Consistency of Bayesian inference for survival analysis with or without censoring
Jayanta K. Ghosh, Ravi Ramamoorthi · Lecture notes-monograph series · 1995
We study the convergence of the posterior distribution to the true distribution in the context of survival analysis data.In the presence of censoring, when the prior is a Dirichlet process, we establish the consistency when the true distribution satisfies a bounded support assumption.We provide a sufficient condition for consistency for general priors.In the uncensored case we prove a similar result when the prior for the survival distribution arises through a Dirichlet Process prior for the hazard rate.