Semiparametric Bayesian Analysis of Survival Data
by
Bani Mallick
Department of Statistics, Texas A & M University, Texas, USA

In medical studies interest centers on the relation of the survival time to the explanatory variables. Bayesian models are considered to develop a regression relationship between survival time and explanatory variables. Modern Statistical work encourages less presumptive; nonparametric specification for at least a portion of the modeling. Various ways this partial nonparametric specification could be incorporated. It can capture the hazard functions, error distributions, transformations. These types of flexible survival models are considered and illustrated using examples with complicated censored survival data. Nonparametric partial exchangable structures are introduced. Markov Chain Monte Carlo algorithms are used to find posterior estimates of several quantities of interest even when we are dealing with complex models and unusual data structures. The work is extended for multivariate survival data.

Date received: November 15, 2000

Copyright © 2000 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cafx-12.