%0 Journal Article
%T Representative of L1/2 Regularization among Lq （0<br>L1/2正则子在Lq（0
%A XU Zong-Ben
%A GUO Hai-Liang
%A WANG Yao
%A ZHANG Hai
%A <br>徐宗本
%A 郭海亮
%A 王尧
%A 张海
%J 自动化学报
%D 2012
%I 
%X Recently, regularization methods have attracted increasing attention. Lq (0 < q < 1) regularizations were proposed after L1 regularization for better solution of sparsity problems. A natural question is which is the best choice among Lq regularizations with all q in (0, 1)? By taking phase diagram studies with a set of experiments implemented on signal recovery and error correction problems, we show the following: 1) As the value of q decreases, the Lq regularization generates sparser solution. 2) When 1/2 ≤ q < 1, the L 1/2 regularization always yields the best sparse solution and when 0 < q ≤ 1/2, the performance of the regularizatons takes no significant difference. Accordingly, we conclude that the L1/2 regularization can be taken as a rep- resentative of Lq (0 < q < 1) regularizations.
%K Lq regularization
%K phase diagram
%K signal recovery
%K error correction<br>Lq
%K regularization
%K phase
%K diagram
%K signal
%K recovery
%K error
%K correction
%U http://www.alljournals.cn/get_abstract_url.aspx?pcid=5B3AB970F71A803DEACDC0559115BFCF0A068CD97DD29835&cid=8240383F08CE46C8B05036380D75B607&jid=E76622685B64B2AA896A7F777B64EB3A&aid=BC10CD3F59E055E28362AC54E10A27EE&yid=99E9153A83D4CB11&vid=16D8618C6164A3ED&iid=DF92D298D3FF1E6E&sid=7D257F36093061DE&eid=323ACE6C9CE37BAB&journal_id=0254-4156&journal_name=自动化学报&referenced_num=0&reference_num=15