l0-norm based structural sparse least square regression for feature selection

作者：

Highlights：

• We impose l0-norm inequality constraint to build the structural sparse LSR problem.

• We develop an adaptive algorithm to ensure the structural sparsity accurately.

• Theoretical results on the efficiency and effectiveness of our method are provided.

• Experimental results prove the superiority of our method over the state-of-the-arts.

摘要

Highlights•We impose l0-norm inequality constraint to build the structural sparse LSR problem.•We develop an adaptive algorithm to ensure the structural sparsity accurately.•Theoretical results on the efficiency and effectiveness of our method are provided.•Experimental results prove the superiority of our method over the state-of-the-arts.

论文关键词：Structural sparse learning,l0-norm,Least square regression,Feature selection,Adaptive greedy algorithm

论文评审过程：Received 22 November 2014, Revised 13 April 2015, Accepted 7 June 2015, Available online 23 June 2015, Version of Record 19 August 2015.

论文官网地址：https://doi.org/10.1016/j.patcog.2015.06.003