THE BENEFIT OF GROUP SPARSITY
THE BENEFIT OF GROUP SPARSITY
复制标题
DOI：
10.1214/09-aos778
复制
发表时间：
2010-08-01
影响因子：
4.5
通讯作者：
Zhang, Tong
中科院分区：
文献类型：
--
作者：
Huang, Junzhou;Zhang, Tong
本文使用强群稀疏性的概念开发了群 Lasso 理论。我们的结果表明，对于强组稀疏信号，组套索优于标准套索。当底层组结构与数据一致时，这为使用组稀疏正则化提供了令人信服的理论依据。此外，该理论还预测了群 Lasso 公式的一些局限性，这些局限性已通过模拟研究得到证实。
This paper develops a theory for group Lasso using a concept called strong group sparsity. Our result shows that group Lasso is superior to standard Lasso for strongly group-sparse signals. This provides a convincing theoretical justification for using group sparse regularization when the underlying group structure is consistent with the data. Moreover, the theory predicts some limitations of the group Lasso formulation that are confirmed by simulation studies.