正则化稀疏模型  

Survey on the Regularized Sparse Models

作　　者：刘建伟[1] 崔立鹏[1] 刘泽宇[2] 罗雄麟[1] 

机构地区：[1]中国石油大学(北京)自动化研究所,北京102249 [2]中国科学院软件研究所基础软件国家工程研究中心,北京100190

出　　处：《计算机学报》2015年第7期1307-1325,共19页Chinese Journal of Computers

基　　金：国家自然科学基金(21006127)资助~~

摘　　要：正则化稀疏模型在机器学习和图像处理等领域发挥着越来越重要的作用,它具有变量选择功能,可以解决建模中的过拟合等问题.Tibshirani提出的Lasso使得正则化稀疏模型真正开始流行.文中总结了各种正则化稀疏模型,指出了各个稀疏模型被提出的原因、所具有的优点、适宜解决的问题及其模型的具体形式.最后,文中还指出了正则化稀疏模型未来的研究方向.The regularized sparse models are playing a more and more important role in many areas,such as in the machine learning and image processing.The regularized sparse models have the ability of variable selection,so they can solve the over-fitting problem.The Lasso proposed by Tibshirani makes regularized sparse models become popular.This paper summarizes various regularized sparse models and points out the motivation of every regularized sparse model,the advantages of every regularized sparse model,the problems that every regularized sparse model can solve and the form of every regularized sparse model.In the end,we offer the regularized sparse models＇ research directions in the future.

关 键 词：正则化 稀疏 变量选择 套索 无偏估计 组稀疏 融合套索 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]