基于雙线性型的非负矩阵集分解.docVIP

PAGE PAGE 9 《计算机学报》2009年8期 基于双线性型的非负矩阵集分解国家自然科学基金项国家自然科学基金项李乐 章毓晋 （清华信息科学与技术国家实验室，北京100084；清华大学电子工程系，北京100084） 摘 要: 非负矩阵分解（non-negative matrix factorization，NMF)是一种常用的非负多元数据描述方法。处理数据矩阵集时，NMF描述力不强、推广性差。为解决这两个问题，并保留NMF的好特性，本文提出了非负矩阵集分解（non-negative matrix set factorization，NMSF)的概念，并在NMSF的框架下系统研究了基于双线性型的非负矩阵集分解（bilinear form-based non-negative matrix set factorization，BFBNMSF)和构造了单调下降的BFBNMSF算法。理论分析和实验结果均表明：处理数据矩阵集时，BFBNMSF比NMF描述力强、推广性好。由此可认为，此时BFBNMSF比NMF更善于抓住数据的本质特征。 关键字：非负矩阵集分解 双线性型 非负矩阵分解 多元数据描述 图像描述 特征提取 Bilinear Form-Based Non-negative Matrix Set Factorization Li Le, Zhang Yu-Jin （Tsinghua National Laboratory for Information Science and Technology, Tsinghua University, Beijing 100084, China; Department of Electronic Engineering, Tsinghua University, Beijing 100084，China） Abstract: Non-negative matrix factorization (NMF) is a popular technique for representations of non- negative multivariate data. While treating a set of matrices, NMF is confronted with two main problems (unsatisfactory accuracy of representation and bad generality). In this paper, non-negative matrix set factorization (NMSF) is conceived to overcome the two problems and to retain NMF’s good properties. Under the frame of NMSF, bilinear form-based non-negative matrix set factorization (BFBNMSF) is systematically studied, and a monotonic algorithm of BFBNMSF is put forward. Theoretical analysis and experimental results show that while processing a data matrix-set, BFBNMSF results in more accurate representation and holds better generality than NMF, therefore it tends to extract more essential features of data matrix sets than NMF. Keywords: non-negative matrix set factorization (NMSF); bilinear form; nonnegative matrix factorization (NMF); multivariate data representation; image representation; feature extraction 1. 引言 在信号处理、神经网络、模式识别、计算机视觉以及图象工程的研究中，如何构造一个能使多维观测数据被更好地描述出来的变换方法始终是一个有基础的重要性的问题。通常，一个好的变换方法应具备两个基本特性：①可以使数据的某种潜在结构变得清晰；②能使数据的维数得到一定程度的约减。 主分量分析、投影寻踪、因子分析、冗余归