On the basis of LINEX asymmetric loss function,the compound LINEX symmetric loss function and the modified squared error loss function,we study the Bayes estimation,E-Bayes estimation and Multi-layer Bayes estimation of Logarithm gamma distribution parameter.There are four chapters in this thesis.In first chapter,the background and research significance of Bayes estimation of scale parameter for log-gamma distribution under different loss function are introduced.In second chapter,the elementary knowledge of Bayes estimation are given.In third chapter,based on a large amount of precise datas,under three kinds of loss functions,we give the Bayes estimation,E-Bayes estimation and Multi-layer Bayes estimation of scale parameter for log-gamma.Finally,through random numerical simulation,their robustness and precision are compared.In fourth chapter,based on fuzzy data,under three kinds of loss functions,we give the Bayes interval estimation and E-Bayes interval estimation of scale parameter for log-gamma distribution.Finally,through random numerical,their interval width are compared. |