Calculating Hazard Function of Survival Model by Bayesian Approach using Linex and General Entropy Loss Function with Jeffrey’s Prior

Abstract

This research observes about the cancer patients in a censored data. The purpose is to work out parameter estimation of hazard function of a survival model. By using Bayesian and giving a prior, posterior formula is constructed from Exponential - Jeffrey’s and Likelihood function. The formula is used to predict the parameter λ under Linex and General Entropy Loss Function (GELF) approach. The parameter estimation of hazard function ĥBL, ĥBGELF and survival function ŝBL, ŝBGELF are determined after and found. The result shows that the value of ĥBL and ĥBGELF are 1,058533880 and 0,008221809, respectively. Thus, we compare the result hazard value under both of Linex and GELF approximation with the real hazard value to find the best method for the research by finding smaller MSE. The MSE of hazard value under Linex and GELF are 1,103155 and 0, respectively. It means that the GELF approach is the best method.