Oracle Properties of the Adaptive Elastic Net

摘要：

This paper proposes an adaptive elastic net for sparse high-dimension regression by incorporating the adaptive penalty mechanism into the elastic net. Under a partial orthogonality condition, the initial univariate regression estimator is shown to be zeroconsistent, based on which the consistent variable selection and the asymptotic estimation normality is obtained. In addition, the controllable group selection capability is also achieved by properly selecting a special reseating factor.

关键词： Statistical learning grouping effect oracle properties

作者: Juntao Li Xuemei Dong Xinlei Li Wenlin Li

作者单位: College of Mathematics and Information Science,henan normal University,XinXiang,453007,China School of Statistics and Mathematics,Zhejiang Gongshang University,Hangzhou,310018,China Henan Normal University,XinXiang,453007,China

会议类型: 国际会议

会议名称: 2010 IEEE International Conference on Intelligent Computing and Intelligent Systems(2010 IEEE 智能计算与智能系统国际会议 ICIS 2010)

会议地点: 厦门

会议语种:英文

页码: 538-542

在线出版日期: 2010-10-29（万方平台首次上网日期，不代表论文的发表时间）