Max-min distance nonnegative matrix factorization

DOI10.1016/J.NEUNET.2014.10.006MaRDI QIDQ889348zbMATH OpenOpenAlexDBLPWikidataFDO

Authors Jim Jing-Yan Wang, Xin Gao

Publication date 6 November 2015

Published in Neural Networks (Search for Journal in Brave)

Copyright license Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International

Full work available at URL https://arxiv.org/abs/1312.1613

zbMATH Keywords

supervised learningnonnegative matrix factorizationdata representationmax-min distance analysis

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Pattern recognition, speech recognition (68T10)

Abstract: Nonnegative Matrix Factorization (NMF) has been a popular representation method for pattern classification problem. It tries to decompose a nonnegative matrix of data samples as the product of a nonnegative basic matrix and a nonnegative coefficient matrix, and the coefficient matrix is used as the new representation. However, traditional NMF methods ignore the class labels of the data samples. In this paper, we proposed a supervised novel NMF algorithm to improve the discriminative ability of the new representation. Using the class labels, we separate all the data sample pairs into within-class pairs and between-class pairs. To improve the discriminate ability of the new NMF representations, we hope that the maximum distance of the within-class pairs in the new NMF space could be minimized, while the minimum distance of the between-class pairs pairs could be maximized. With this criterion, we construct an objective function and optimize it with regard to basic and coefficient matrices and slack variables alternatively, resulting in a iterative algorithm.

Recommendations

• Class-driven non-negative matrix factorization for image representation
• Evaluation of distance metrics for recognition based on non-negative matrix factorization
• Hyperplane-based nonnegative matrix factorization with label information
• Image Classification with Nonnegative Matrix Factorization Based on Spectral Projected Gradient
• Nonnegative matrix factorization with rank regularization and hard constraint

Cites work

• scientific article; zbMATH DE number 3518555 (Why is no real title available?)
• A comparative study of ordinary cross-validation, v-fold cross-validation and the repeated learning-testing methods
• Alternative optimization strategies for large-scale production-allocation problems
• Duality theory for optimization problems with interval-valued objective functions
• Generalized Lagrange function and generalized weak saddle points for a class of multiobjective fractional optimal control problems
• Introducing a weighted non-negative matrix factorization for image classification
• Learning the parts of objects by non-negative matrix factorization
• Multiple graph regularized nonnegative matrix factorization
• Non-negative matrix factorization with sparseness constraints
• Nonnegative matrix factorization and its applications in pattern recognition
• Optimality and duality theory for stochastic optimization problems with nonlinear dominance constraints
• Projected Gradient Methods for Nonnegative Matrix Factorization

Cited in

(4)

• Inter-class sparsity based discriminative least square regression
• scientific article; zbMATH DE number 1203228 (Why is no real title available?)
• Orthogonal self-guided similarity preserving projection for classification and clustering
• Class-driven non-negative matrix factorization for image representation

Describes a project that uses

Uses Software

• ESLpred
• UniProt
• Yale Face

This page was built for publication: Max-min distance nonnegative matrix factorization

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q889348)