Metode Bayesian untuk Estimasi Parameter Distribusi Eksponensial pada Data Tersensor
DOI:
https://doi.org/10.21009/jmt.4.2.3Keywords:
censored data, Bayesian parameter estimation, conjugate prior, SELF, survival analysisAbstract
Parameter is a value that describe the characteristics of a population. But the parameter of a real data, the value is unknown. To estimate the value of the parameter, there are several methods, which are maximum likelihood estimation method (MLE) and Bayesian parameter estimation method. In Bayesian method, the prior information is applied to update the current data. The prior is determined based on the information in the data. This article using censored data with exponential distribution, and using the conjugate prior. Followed by squared error loss function (SELF), the estimated value function on the λ parameter. When the function was applied on Stanford heart transplant data, the value of ˆλ = 0.00089, which means the patient’s failure (death) probability is low and the patient’s probability to survive is high.