摘要
Lasso方法是变量选择领域的一类重要方法,它具有计算上的简便性.SCAD和MCP在一定程度上降低了Lasso估计的偏差.本文在对影响我国各省房价的主要因素进行数据分析时,对比研究了Lasso,SCAD和MCP的效果。
The Lasso method is an important method in the field of variable selection with computational simplicity.SCAD and MCP reduce the errors of Lasso estimation to a certain extent.In the paper below,a comparison was made of the effects of Lasso,SCAD and MCP in an analysis of the main factors affecting housing prices in various provinces of China.
作者
王璐
孙聚波
WANG Lu;SUN Ju-bo(Jilin Engineering Normal University,Changchun Jilin 130052,China)
出处
《吉林工程技术师范学院学报》
2021年第12期109-112,共4页
Journal of Jilin Engineering Normal University
