摘要
通过比较最小二乘回归模型和稳健回归模型在民营经济发展水平与自然和社会属性相关性的研究,发现稳健回归相较于最小二乘估计可以有效减小异常值对回归方程的影响,能更好的揭示发展水平的相关性因素。文末给出针对性政策建议。
By comparing the least squares regression model with the robust regression model on the study of correlation between the private economic development level and the natural and social attributes,it is found that the robust regression can effectively reduce the influences of outliers compared to the least square estimates of regression functions,and can reveal the relevant factors of the level of development better.And some policy recommendations are given.
作者
杨联强
叶梦琳
程丹丹
刘美月
YANG Lian-qiang;YE Meng-lin;CHENG Dan-dan;LIU Mei-yue(School of Mathematical Science,Anhui University,Hefei 230601,China)
出处
《合肥学院学报(综合版)》
2022年第2期32-37,共6页
Journal of Hefei University:Comprehensive ED
基金
安徽大学创新创业训练计划“稳健回归模型理论及其在我国上市公司区域分布研究中的应用”(Y020412012/257)
安徽省高校自然科学基金“机器学习中稳健众值回归模型”(KJ2021A0049)资助。
关键词
民营经济
相关性
稳健回归
private economy
correlation
robust regression