Activity Number:
|
370
|
Type:
|
Invited
|
Date/Time:
|
Thursday, August 15, 2002 : 8:30 AM to 10:20 AM
|
Sponsor:
|
WNAR
|
Abstract - #300380 |
Title:
|
Bayesian Non-parametric Modeling of Proportional Hazards and Frailty Models
|
Author(s):
|
Pamela Ohman Strickland*+ and Richard Levine and Juanjuan Fan
|
Affiliation(s):
|
University of Medicine and Dentistry of New Jersey and University of California, Davis and University of California, Davis
|
Address:
|
Liberty Plaza, Suite 2200, P.O. Box 2688, New Brunswick, NJ, 08903-2688,
|
Keywords:
|
|
Abstract:
|
Bayesian nonparametric modeling is examined in the context of the analysis of both univariate and clustered time-to-event data. Assuming the Cox proportional hazards framework, we examine the implication of Bayesian nonparametric modeling of, first, the baseline hazard and, second, for clustered data the distribution of random effects, which have a multiplicative effect on the hazard. The Bayesian nonparametric methods can be helpful in smoothing the hazard function and describing the heterogeneity of the population. We also examine the implications on estimation and diagnostics. Advantages and disadvantages of using the non-parametric approach are examined, with emphasis on the practical implications rather than technical details.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2002 program |