《The Benefit of Group Sparsity》

Junzhou Huang and Tong Zhang，"The Benefit of Group Sparsity", Ann. Statist. Volume 38, Number 4 (2010), 1978-2004.

作者从理论上了分析了group lasso和lasso的性能情况，分析了在何种条件下group lasso可以得到更好的结果。

Group Lasso is more robust to noise due to the stability associated with group structure.

Group Lasso requires a smaller sample size to satisfy the sparse eigenvalue condition required
in the modern sparsity analysis.

However, group Lasso can be inferior if the signal is only weakly group-sparse, or covered by groups
with small sizes. Moreover, group Lasso does not perform well with overlapping groups (which is
not analyzed in this paper).

Group lasso是通过组内系数的L2范数来描述系数向量的组结构，我考虑的是通过组内系数与组内样本共同来描述这种组结构。

转载于:https://www.cnblogs.com/bingxin-xu/p/3358254.html