|
Activity Number:
|
66
- Frontiers in Bayesian Nonparametric Methods, from Theory to Application
|
|
Type:
|
Invited
|
|
Date/Time:
|
Monday, August 9, 2021 : 10:00 AM to 11:50 AM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
|
Abstract #316998
|
|
|
Title:
|
A Bayesian Nonparametric Model via Mixtures of Erlang Distributions for Survival Times
|
|
Author(s):
|
Juhee Lee*
|
|
Companies:
|
University of California, Santa Cruz
|
|
Keywords:
|
|
|
Abstract:
|
We propose Bayesian nonparameric modeling for survival responses built from mixtures of Erlang densities, which offer a flexible, albeit parsimonious mixture representation for the survival density and the hazard function. The mixture weights are defined through increments of a distribution function modeled with a Dirichlet process prior. This construction enables extensions to flexible regression modeling. The proposed models have exciting potential with regard to their inferential and predictive capacity, and they also offer a new perspective relative to state-of-the-art methods. Performance of the proposed model is examined by simulation, and the model is applied to analyze a real survival dataset, including comparisons to a conventional proportional hazards regression model.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2021 program
|
