A Quasi-Likelihood Approach to Nonnegative Matrix Factorization

作     者：Devarajan, Karthik Cheung, Vincent C. K. 

作者机构：Temple Univ Hlth Syst Dept Biostat & Bioinformat Fox Chase Canc Ctr Philadelphia PA 19111 USA Chinese Univ Hong Kong Sch Biomed Sci Shatin Hong Kong Peoples R China 

出 版 物：《NEURAL COMPUTATION》 (神经计算)

年 卷 期：2016年第28卷第8期

页      面：1663-1693页

核心收录：

学科分类：1001[医学-基础医学(可授医学、理学学位)] 08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：NIH [P30 CA06927] Chinese University of Hong Kong 

主　　题：expectation-maximisation algorithm factorization statistical models (nuclear) Nonnegative matrix decomposition family Perspective goodness of fit linear models Frameworks medical signal processing Converge 

摘      要：A unified approach to nonnegative matrix factorization based on the theory of generalized linear models is proposed. This approach embeds a variety of statistical models, including the exponential family, within a single theoretical framework and provides a unified view of such factorizations from the perspective of quasi-likelihood. Using this framework, a family of algorithms for handling signal-dependent noise is developed and its convergence proved using the expectation-maximization algorithm. In addition, a measure to evaluate the goodness of fit of the resulting factorization is described. The proposed methods allow modeling of nonlinear effects using appropriate link functions and are illustrated using an application in biomedical signal processing.