单指标众数模型的Bayes局部影响分析  

Bayesian Local Influence Analysis and Empirical Study of Single Index Mode Model

作　　者：朱桂玲 ZHU Guiling(School of Mathematic and Statistics,Zhaotong University,Zhaotong Yunnan 657000,China)

机构地区：[1]昭通学院数学与统计学院,云南昭通657000

出　　处：《保山学院学报》2023年第5期43-47,共5页JOURNAL OF BAOSHAN UNIVERSITY

基　　金：云南省教育厅科学研究基金项目“缺失数据下高维数据的假设检验”(项目编号:2016FB005)。

摘　　要：近年来,单指标众数模型基于数据删除模型和众数漂移模型的统计诊断与局部影响分析深受广大学者的青睐。这些都是从频率的角度基于该模型利用一些诊断统计量来进行数据分析。从Bayes的角度基于MCMC方法研究单指标众数模型的局部影响分析,得到因变量和自变量的扰动结果,并根据扰动结果判断异常值。In recent years,statistical diagnosis and local impact analysis of single index mode model based on data deletion model and mode drift model have been favored by many scholars.These are all from a frequency perspective based on the model using some diagnostic statistics to analyze the data.This paper will study the local impact analysis of the single index mode model from the perspective of Bayes based on MCMC method,obtain the disturbance results of dependent variables and independent variables,and judge the outliers according to the disturbance results.

关 键 词：单指标众数模型 MCMC方法 局部影响分析 BAYES分析 

分 类 号：O2[理学—数学]