| 基于分布估计算法的非负矩阵分解 |
| |
| 引用本文： | 孟佳音,刘丁酉. 基于分布估计算法的非负矩阵分解[J]. 湖北民族学院学报(自然科学版), 2012, 0(3): 241-244 |
| |
| 作者姓名： | 孟佳音 刘丁酉 |
| |
| 作者单位： | 北京市十一学校;武汉大学数学与统计学院 |
| |
| 基金项目： | 国家自然科学基金项目(60825007) |
| |
| 摘 要： | 非负矩阵分解问题可以转化为一个约束优化问题,因此可以依靠最优化领域的相关算法进行求解.提出一种基于分布估计算法求解非负矩阵分解问题的新算法,并将算法应用于两个非负矩阵分解的数值算例,与非负矩阵分解基准算法进行比较,证实了算法的可行性和优越性.
|
| 关 键 词： | 非负矩阵分解 分布估计算法 最小二乘法 |
Non-negative Matrix Factorization Based on Estimation of Distribution Algorithm |
| |
| MENG Jia-yin,LIU Ding-you. Non-negative Matrix Factorization Based on Estimation of Distribution Algorithm[J]. Journal of Hubei Institute for Nationalities(Natural Sciences), 2012, 0(3): 241-244 |
| |
| Authors: | MENG Jia-yin LIU Ding-you |
| |
| Affiliation: | (2,*) (1.Beijing National Day School,Beijing 100039,China; 2.School of Mathematics and Statistics,Wuhan University,Wuhan 430072,China) |
| |
| Abstract: | Since non-negative matrix factorization problem(NMF) can be developed into a constrained optimization problem,we can use optimization algorithm to handle this problem.This paper proposes a new NMF algorithm based on estimation of distribution algorithm(EDA),and applies the proposed algorithm to solve two numerical problems.The results are compared with those obtained by the basic non-negative matrix factorization algorithm,and the comparison reveals that the effectiveness and efficiency of the proposed algorithm. |
| |
| Keywords: | non-negative matrix factorization estimation of distribution algorithm method of least square |
| 本文献已被 CNKI 等数据库收录！ |
|
