Non-convex Optimization Problem in Machine Learning and Data Mining

• 作 者: ZhangChangshui;
• 机 构: Tsinghua Univ.
• 出 处: 第八届工业与应用数学国际大会
• 发表日期: 2015年期

　　In machine and data mining,optimization is a very important problem.In this talk,we introduce optimization problems in machine learning,and describe three non-convex optimization algorithms: an algorithm for decomposable non-convex regularization,an algorithm for feature selection in multi-task learning,and an adaptive algorithm for non-convex regularization.