Sparse Representation (SR) Toolbox
Application Background: This toolbox implements machine learning methodologies including sparse coding-based classification, dictionary-based dimensionality reduction with sub-dictionary learning, learning models, and linear regression/classification (LRC). It features implementations of kernel l1-regularized and/or non-negative constrained sparse coding and dictionary learning models. Key Technologies: The optimization utilizes active set, interior point, proximal, and decomposition methods. Current version: 1.9 (March 2, 2015). Freely available for academic use with commercial licenses offering advanced features and technical support.