摘要
残差分析是回归诊断中的重要内容。文章研究了残差向量的相关性质,梳理了残差向量的分类,并通过对经典的回归数据进行扰动,对比分析几种残差在异常点检测中的作用,为进一步诊断提供依据。
Residual analysis is an important part of regression diagnosis. This paper studies the related properties of residual vectors, sorts out the classification of residual vectors, and compares and analyzes the role of several residual errors in anomaly detection through perturbation of classical regression data, so as to provide a basis for further diagnosis.
作者
曹桃云
卢利敏
延丽平
Cao Taoyun;Lu Limin;Yan Liping(Big Data and Education Statistics Application Laboratory,Guangdong University of Finanoe and Economics,Guangzhou 510320,China;School of Statistics and Mathematics,Guangdong University of Finanoe and Economics,Guangzhou 510320,China;Information Engineering School,Guangzhou Vocational College of Technology and Business,Guangzhou 511442,China)
出处
《统计与决策》
CSSCI
北大核心
2022年第12期28-31,共4页
Statistics & Decision
基金
广东省自然科学基金面上项目(2020A1515011580)
广东财经大学校级学位与研究生教育改革研究项目(2021YB08)
广东高校省级重点平台和重大科研项目特色创新项目(2018GKTSCX010)。
关键词
回归诊断
普通残差
内学生化残差
外学生化残差
regression diagnosis
ordinary residual
internal studentized residuals
external studentized residuals
