非线性均值方差模型的贝叶斯数据删除统计诊断  

Bayesian Case Deletion Statistical Diagnostics for Nonlinear Mean and Variance Models

作　　者：赵远英[1] 徐登可[2] 段星德 ZHAO Yuanying;XU Dengke;DUAN Xingde(College of Mathematics and Information Science,Guiyang University,Guiyang 550005;Department of Statistics,Zhejiang Agriculture and Forestry University,Hangzhou 311300;School of Mathematics and Statistics,Guizhou University of Finance and Economy,Guiyang 550025)

机构地区：[1]贵阳学院数学与信息科学学院,贵阳550005 [2]浙江农林大学统计系,杭州311300 [3]贵州财经大学数学与统计学院,贵阳550025

出　　处：《系统科学与数学》2020年第1期171-179,共9页Journal of Systems Science and Mathematical Sciences

基　　金：国家自然科学基金项目(11761016,11501073,11801514);浙江省自然科学基金项目(LY17A010026);贵州省科学技术基金项目(黔科合LH字[2014]7164号,黔科合LH字[2015]7299号);贵阳市科技局贵阳学院专项资金项目(GYU-KYZ[2018]04);贵州省高等学校大学生创新创业训练计划项目(109762018520837,109762018520860);贵阳学院2018年课程建设资助项目(概率与统计)资助课题。

摘　　要：文章在非线性均值方差模型框架下基于K-L距离研究贝叶斯数据删除影响的统计诊断问题,通过应用Gibbs抽样和MH算法估计贝叶斯数据删除影响诊断统计量.随机模拟研究和红鳟鲑鱼数据的数值例子说明该诊断方法的可行性.Bayesian case deletion statistical diagnostics based on K-L divergence for nonlinear mean and variance models are studied in this paper,in which Gibbs sampler and Metropolis-Hastings algorithm are employed to calculate Bayesian case deletion statistic.Simulation studies and a sockeye salmon data example are used to demonstrate the proposed methodology.

关 键 词：贝叶斯数据删除统计量 Gibbs抽样 K-L距离 MH算法 非线性均值方差模型 

分 类 号：O212.1[理学—概率论与数理统计]