Nonparametric Bayesian modeling for baseline hazard functions†
Yena Jeon1 · Sejoong Kim2 · Jang-Hee Cho3 · Yongku Kim4
14Department of Statistics, Kyungpook National University
2Department of Internal Medicine, Seoul National University Bundang Hospital
3Department of Internal Medicine, Kyungpook National University Hospital
2Department of Internal Medicine, Seoul National University Bundang Hospital
3Department of Internal Medicine, Kyungpook National University Hospital
Correspondence to: † This research was supported by a grant of Patient-Centered Clinical Research Coordinating Center (PACEN) funded by the Ministry of Health & Welfare, Republic of Korea (HI19C0481, HC20C0054).
1 Graduate Student, Department of Statistics, Kyungpook National University, Daegu 41566, Korea.
2 Professor, Department of Internal Medicine, Seoul National University Bundang Hospital, Seongnam 13620, Korea.
3 Professor, Department of Internal Medicine, Kyungpook National University Hospital, School of Medicine, Daegu 41944, Korea.
4 Professor, Department of Statistics, Kyungpook National University, Daegu 41566, Korea. E-mail: kim.1252@knu.ac.kr
1 Graduate Student, Department of Statistics, Kyungpook National University, Daegu 41566, Korea.
2 Professor, Department of Internal Medicine, Seoul National University Bundang Hospital, Seongnam 13620, Korea.
3 Professor, Department of Internal Medicine, Kyungpook National University Hospital, School of Medicine, Daegu 41944, Korea.
4 Professor, Department of Statistics, Kyungpook National University, Daegu 41566, Korea. E-mail: kim.1252@knu.ac.kr
Received June 7, 2022; Revised June 22, 2022; Accepted June 24, 2022.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
- Abstract
- Survival analysis is primarily used to identify the time-to-event for events of interest. The Cox proportional hazards model takes advantage that it accounts for the proportionate risks of covariates without estimating exact baseline hazards. However, estimation of exact hazard distribution is always accompanied by estimating baseline hazard function as well as regression parameters. In this study, we adopted nonparametric Bayesian hierarchical model with exible priors in estimating cumulative base- line hazard function. We assume a monotone step function for the cumulative baseline hazard function, where the number, size, and location of jumps are random. By Estimating the step function through stick-breaking construction, we can obtain a totally data-driven step function.
- Keywords : Baseline hazard function, Cox proportional hazard regression, nonparametric Bayesian analysis, survival analysis.
Full Text(PDF) Free
