Correntropy-Based document clustering via nonnegative matrix factorization

Authors:
Tolga Ensari;Jan Chorowski;Jacek M. Zurada
Affiliations:
Computer Engineering Department, Istanbul University, Istanbul, Turkey;Electrical and Computer Engineering, University of Louisville, Louisville;Electrical and Computer Engineering, University of Louisville, Louisville
Venue:
ICANN'12 Proceedings of the 22nd international conference on Artificial Neural Networks and Machine Learning - Volume Part II
Year:
2012

Citing 13
Cited 0

Introducing a weighted non-negative matrix factorization for image classification

Pattern Recognition Letters
Non-negative Matrix Factorization with Sparseness Constraints

The Journal of Machine Learning Research
Introduction to Data Mining, (First Edition)

Introduction to Data Mining, (First Edition)
Projected Gradient Methods for Nonnegative Matrix Factorization

Neural Computation
Enhancing the correntropy MACE filter with random projections

Neurocomputing
Nonnegative matrix factorization with the itakura-saito divergence: With application to music analysis

Neural Computation
Bayesian Non-negative Matrix Factorization

ICA '09 Proceedings of the 8th International Conference on Independent Component Analysis and Signal Separation
Using correntropy as a cost function in linear adaptive filters

IJCNN'09 Proceedings of the 2009 international joint conference on Neural Networks
Document clustering using nonnegative matrix factorization

Information Processing and Management: an International Journal
Maximum Correntropy Criterion for Robust Face Recognition

IEEE Transactions on Pattern Analysis and Machine Intelligence
Algorithms for nonnegative matrix factorization with the β-divergence

Neural Computation
Correntropy: Properties and Applications in Non-Gaussian Signal Processing

IEEE Transactions on Signal Processing
Robust Principal Component Analysis Based on Maximum Correntropy Criterion

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Nonnegative Matrix Factorization (NMF) is one of the popular techniques to reduce the number of attributes of the data. It has been also widely used for clustering. Several types of the objective functions have been used for NMF in the literature. In this paper, we propose to maximize the correntropy similarity measure to produce the factorization itself. Correntropy is an entropy-based criterion defined as a nonlinear similarity measure. Following the discussion of minimization of the correntropy function, we use it to cluster document data set and compare its clustering performance with the Euclidean Distance (EucD)-based NMF. The comparison is illustrated with 20-Newsgroups data set. The results show that our approach has better clustering compared with other methods which use EucD as an objective function.