Open AccessThis article is
- freely available
Generalized Alpha-Beta Divergences and Their Application to Robust Nonnegative Matrix Factorization
Laboratory for Advanced Brain Signal Processing, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako, 351-0198 Saitama, Japan
Systems Research Institute, Intelligent Systems Laboratory, PAS, Newelska 6 str., 01-447 Warsaw, Poland
Dpto de Teoría de la Señal y Comunicaciones, University of Seville, Camino de los Descubrimientos s/n, 41092-Seville, Spain
Laboratory for Mathematical Neuroscience, RIKEN BSI, Wako, 351-0198 Saitama, Japan
* Authors to whom correspondence should be addressed.
Received: 13 December 2010; in revised form: 4 January 2011 / Accepted: 4 January 2011 / Published: 14 January 2011
Abstract: We propose a class of multiplicative algorithms for Nonnegative Matrix Factorization (NMF) which are robust with respect to noise and outliers. To achieve this, we formulate a new family generalized divergences referred to as the Alpha-Beta-divergences (AB-divergences), which are parameterized by the two tuning parameters, alpha and beta, and smoothly connect the fundamental Alpha-, Beta- and Gamma-divergences. By adjusting these tuning parameters, we show that a wide range of standard and new divergences can be obtained. The corresponding learning algorithms for NMF are shown to integrate and generalize many existing ones, including the Lee-Seung, ISRA (Image Space Reconstruction Algorithm), EMML (Expectation Maximization Maximum Likelihood), Alpha-NMF, and Beta-NMF. Owing to more degrees of freedom in tuning the parameters, the proposed family of AB-multiplicative NMF algorithms is shown to improve robustness with respect to noise and outliers. The analysis illuminates the links of between AB-divergence and other divergences, especially Gamma- and Itakura-Saito divergences.
Keywords: nonnegative matrix factorization (NMF); robust multiplicative NMF algorithms; similarity measures; generalized divergences; Alpha-; Beta-; Gamma- divergences; extended Itakura-Saito like divergences; generalized Kullback-Leibler divergence
Article StatisticsClick here to load and display the download statistics.
Notes: Multiple requests from the same IP address are counted as one view.
Cite This Article
MDPI and ACS Style
Cichocki, A.; Cruces, S.; Amari, S.-I. Generalized Alpha-Beta Divergences and Their Application to Robust Nonnegative Matrix Factorization. Entropy 2011, 13, 134-170.
Cichocki A, Cruces S, Amari S-I. Generalized Alpha-Beta Divergences and Their Application to Robust Nonnegative Matrix Factorization. Entropy. 2011; 13(1):134-170.
Cichocki, Andrzej; Cruces, Sergio; Amari, Shun-ichi. 2011. "Generalized Alpha-Beta Divergences and Their Application to Robust Nonnegative Matrix Factorization." Entropy 13, no. 1: 134-170.