本文已被：浏览 2361次 下载 40次
投稿时间：2014-06-11 修订日期：2014-11-07
投稿时间：2014-06-11 修订日期：2014-11-07
中文摘要: 针对非负矩阵分解中系数矩阵不够稀疏的问题，提出一个新的约束非负矩阵分解算法。在经典非负矩阵分解的优化函数中施加稀疏性约束，并对分解系数矩阵施加最小相关约束，与此同时对基矩阵施加2-范数约束，在保证非负约束和分解精度的基础上，使分解后得到的矩阵尽可能稀疏，这样可以更加节省存储空间，分解结果更优。对比实验表明,提出的算法具有更好的稀疏性，且实验误差更小。
中文关键词: 非负矩阵分解（NMF） 稀疏性 最小相关系数 2-范数
Abstract:Aiming at the lack of sparseness of factorization matrix in the nonnegative matrix factorization (NMF) algorithm,a new constrained NMF algorithm was proposed.A sparseness constraint was added to the original nonnegative matrix factorization (NMF) algorithm,and the minimum correlation constraint was imposed on the coefficient matrices and the 2-norm constraint was imposed on the basis matrix at the same time,which can ensure the non-negative constraint and accurate decomposition,and can also make the decomposed matrix sparse as far as possible,saving more storage space.Comparison with the experiments showed that the propose algorithm has the better sparseness and smaller error than both the original NMF algorithm and the SNMF algorithm.
keywords: non-negative matrix factorization (NMF) sparseness the least correlated component constraints 2-norm
文章编号：201400619 中图分类号： 文献标志码：
基金项目:河海大学中央高校基本科研业务费项目(2013B32514)；国家自然科学基金资助项目(61101211)
| Author Name | Affiliation |
| Li Chenming | Collage of Computer and Info. Eng.,Hohai Univ. |
| Zhang Shiming | Collage of Computer and Info. Eng.,Hohai Univ. |
| Li Changli | Collage of Computer and Info. Eng.,Hohai Univ. |
作者简介:
引用文本：
李臣明,张师明,李昌利.非负矩阵分解的一个约束稀疏算法[J].工程科学与技术,2015,47(2):108-111.
Li Chenming,Zhang Shiming,Li Changli.A Constrained Sparse Algorithm for Nonnegative Matrix Factorization[J].Advanced Engineering Sciences,2015,47(2):108-111.
引用文本：
李臣明,张师明,李昌利.非负矩阵分解的一个约束稀疏算法[J].工程科学与技术,2015,47(2):108-111.
Li Chenming,Zhang Shiming,Li Changli.A Constrained Sparse Algorithm for Nonnegative Matrix Factorization[J].Advanced Engineering Sciences,2015,47(2):108-111.