Bayes estimation of a two-parameter exponential distribution and its implementation

Abstract

Life test data analysis is a statistical method used to analyze time data until a certain event occurs. If the life test data is produced after the experiment has been running for a set amount of time, the life time data may be type I censored data. When conducting observations for survival analysis, it is anticipated that the data would conform to a specific probability distribution. Meanwhile, to determine the characteristics of a population, parameter estimation is carried out. The purpose of this study is to use the linear exponential loss function method to derive parameter estimators from the exponential distribution of two parameters on type I censored data. The prior distribution used is a non-informative prior with the determination technique using the Jeffrey’s method. Based on the research results that have been obtained, application is carried out on real data. This data is data on the length of time employees have worked before they experienced attrition with a censorship limit based on age, namely 58 years, obtained from the Kaggle.com website. Based on the estimation results, the average length of work for employees is 6.29427 years. This shows that employees tend to experience attrition after working for a relatively long period of time.