Bayesian Competing Risk Model for Medical Data
- DOI
- 10.2991/aer.k.211222.003How to use a DOI?
- Keywords
- Bayesian approach; flat prior; gestational time; MCMC; survival analysis
- Abstract
Problems considering group assignment is often found in health area, where the groups represent whether a patient will be recovered or not, at high risk of relapse or not, and many other. While the occurrence of these events could be modelled using classification methods, more insights on the time of occurrence is cannot be provided. Thus, a more comprehensive method is required, which could be answered by the survival model. Competing risk model is one of the statistical methods that could be used for modelling the occurrence of several competing events, in which we can produce not just the probability of the events occurrence, but more specifically the probability of that events at a given time period. In this study, we propose the use of Bayesian competing risk model to predict whether a patient will give birth under the complication of Pre-Eclampsia (PE) condition, for a given gestational time. Data on patients in their first trimester of pregnancy from one of the hospitals in Jakarta were used in the analysis. Non-informative prior distributions were set for the parameters, data were assumed to follow a Weibull distribution, and upon obtaining the posterior distribution, Markov Chain Monte Carlo (MCMC) was implemented for posterior sampling. The result showed fast convergence, as only 30,000 iterations are required to achieve it, and several important predictors were identified.
- Copyright
- © 2021 The Authors. Published by Atlantis Press International B.V.
- Open Access
- This is an open access article under the CC BY-NC license.
Cite this article
TY - CONF AU - Nadya Devana AU - Sarini Abdullah PY - 2021 DA - 2021/12/23 TI - Bayesian Competing Risk Model for Medical Data BT - Proceedings of the International Conference on Science and Engineering (ICSE-UIN-SUKA 2021) PB - Atlantis Press SP - 12 EP - 19 SN - 2352-5401 UR - https://doi.org/10.2991/aer.k.211222.003 DO - 10.2991/aer.k.211222.003 ID - Devana2021 ER -