[1]刘志扬.非负矩阵分解及其改进方法[J].华侨大学学报(自然科学版),2016,37(6):782-785.[doi:10.11830/ISSN.1000-5013.201606025]
　LIU Zhiyang.Research on Non Negative Matrix Factorization and It’s Improvement Method[J].Journal of Huaqiao University(Natural Science),2016,37(6):782-785.[doi:10.11830/ISSN.1000-5013.201606025]
点击复制

非负矩阵分解及其改进方法()
分享到：

《华侨大学学报（自然科学版）》[ISSN:1000-5013/CN:35-1079/N]

卷:
第37卷
期数:
2016年第6期
页码:
782-785
栏目:
出版日期:
2016-11-20

文章信息/Info

Title:
Research on Non Negative Matrix Factorization and It’s Improvement Method
文章编号:
1000-5013(2016)06-0782-04
作者:
刘志扬
广东科技学院 基础部, 广东 东莞 523083
Author(s):
LIU Zhiyang
Basic Course Department, Guangdong University of Science and Technology, Dongguan 523083, China
关键词:
非负矩阵;  非负分解;  优化函数;  迭代方程
Keywords:
non negative matrix;  non negative decomposition;  optimization function;  iterative equation
分类号:
O151.21
DOI:
10.11830/ISSN.1000-5013.201606025
文献标志码:
A
摘要:
首先,给出非负矩阵分解的数学形式,分析欧式距离和相对熵(KL)散度两种分解误差评价函数.然后,针对3种特殊形式的非负矩阵进行分解方法的改进,优化函数和迭代过程分别适用于正交非负矩阵、凸非负矩阵、投影非负矩阵的分解.结果表明:提出的改进方法简化了非负矩阵分解的过程.
Abstract:
First of all, mathematical form of non negative matrix factorization is presented, and two decomposition error evaluation functions of Euclidean distance and relative entropy divergence are presented. Then, we improve the decomposition method for 3 kinds of non negative matrix, the used optimization function and the iteration process can respectively applied to the decomposition of orthogonal nonnegative matrix, convex nonnegative matrix and projection non negative matrix. The results show that the improved method simplifies the process of non negative matrix factorization.

参考文献/References:

[1] 李磊,李静.三端点区间数的判断矩阵分解及基于Fermat的算法集结研究[J].数学的实践与认识,2015,45(17):214-221.
[2] NEDUNGADI P,SMRUTHY T K.Personalized multi-relational matrix factorization model for predicting student performance[J].Advances in Intelligent Systems and Computing,2016,38(4):163-172.
[3] 方蔚涛,马鹏,成正斌,等.二维投影非负矩阵分解算法及其在人脸识别中的应用[J].自动化学报,2012,38(9):1503-1512.
[4] 涂丹丹,舒承椿,余海燕.基于联合概率矩阵分解的上下文广告推荐算法[J].软件学报,2013,24(3):454-464.
[5] MESAROS A,HEITTOLA T,DIKMEN O,et al.Sound event detection in real life recordings using coupled matrix factorization of spectral representations and class activity annotations[C]//International Conference on Acoustics, Speech and Signal Proceeding.London:AEAT,2015:151-155.
[6] 刘维湘,郑南宁,游屈波.非负矩阵分解及其在模式识别中的应用[J].科学通报,2006,51(3):241-250.
[7] 杨粤涛,朱明,贺柏根,等.采用改进投影梯度非负矩阵分解和非采样Contourlet变换的图像融合方法[J].光学精密工程,2011,19(5):1143-1151.
[8] 宋海州,徐强,田朝薇.计算非负不可约矩阵谱半径的新算法[J].华侨大学学报(自然科学版),2011,32(3):348-351.
[9] KITAMURA D,ONO N,SAWADA H,et al. Efficient multichannel nonnegative matrix factorization exploiting rank-1spatial model[C]//International Conference on Acoustics, Speech and Signal Proceeding.London:AEAT,2015:276-280.
[10] LUDENA C J,GALLARDO A A.Acoustic event classification using spectral band selection and non-negative matrix factorization-based features[J].Expert Systems with Applications,2016,46(8):77-86.

备注/Memo

备注/Memo:
收稿日期: 2016-10-13
通信作者: 刘志扬(1964-),男,副教授,主要从事非负矩阵分解算法的研究.E-mail:nbxylzy@163.com.
基金项目: 广东省教育厅、财政厅立项资助课题(2013WYXM0136)
更新日期/Last Update: 2016-11-20