讲座主题: Nearly Efficient Closed-form Estimators for the Beta Distribution
主讲人：陈飘
讲座时间：2021-5-20（周四）10:30-11:30

地点：综合楼644

主讲人简介：    

 Dr. Piao Chen is an assistant professor in statistics at Delft Institute of Applied Mathematics, Delft University of Technology. He obtained his PhD in Industrial and Systems Engineering Management from National University of Singapore in 2017, and Bachelor in Industrial Engineering from Shanghai Jiao Tong University in 2013. His current research interests include developing inference methods for common probability distributions and stochastic processes, modelling and outbreak detection of infectious diseases, and reliability engineering including survival analysis and degradation modelling. His work has appeared in journals including Technometrics, Journal of Quality Technology, Statistics in Medicine, Naval Research Logistics, IISE Transactions, European Journal of Operational Research, IEEE Transactions on Reliability, and Reliability Engineering & System Safety.、

讲座摘要：

The beta distribution is commonly used to model random variables restricted in the finite intervals. In practice, the maximum likelihood (ML) estimation and the method of moment are often used to estimate the two parameters of the beta distribution. Because the ML estimators do not have closed forms and the moment estimators are not efficient, however, these classical estimators are not appropriate or even not available in certain applications. In view of this fact, novel closed-form estimators of the beta distribution are proposed in this study. The underlying idea is to involve the log-moments in the estimation equations and solve the mixed type of moment equations simultaneously. The resulted estimators are in closed forms, strongly consistent and nearly efficient. In addition, through extensive simulations, the proposed estimators are shown to perform almost identically to the ML estimators in both small and large samples, and they significantly outperform the moment estimators. Given the systematic biases of the proposed estimators in finite samples, the closed-form bias-corrected estimators are further proposed, and their outstanding performance is verified by simulations.