矩陣分解(MF)方法及程式碼

風景不在對岸wj發表於2017-09-21

一、基於投影梯度法的非負矩陣分解

論文:Projected gradient methods for non-negative matrix factorization 
程式碼:Matlab及Python原始碼

二、基於類牛頓法的最小二乘矩陣近似解法

論文:Fast Newton-type Methods for the Least Squares Nonnegative Matrix Approximation Problem 
PPT:https://www.niss.org/sites/default/files/Dhillon_workshop.pdf

三、基於alternating nonnegative least squares(ANLS)框架的一種NMF分解方法

論文:Toward Faster Nonnegative Matrix Factorization: A New Algorithm and Comparisons

四、 active-set-like演算法求解NMF

論文: Fast nonnegative matrix factorization: An active-set-like method and comparisons



原文地址

相關文章