This package reproduces results from
Robert Peharz and Franz Pernkopf, "Sparse Nonnegative Matrix Factorization with l0-constraints", Neurocomputing, vol. 80, pp. 38--46, 2012.
In particular, it provides algorithms for approximate non-negative matrix factorization with l0-sparseness constraints.
PLEASE NOTE THE ACCOMPANYING LICENSE FILE (modified BSD, 3-Clause). IF YOU USE THIS CODE FOR RESEARCH, PLEASE CITE THE PAPER ABOVE.
Overview:
NMFL0_H.m: implements approximate NMF with l0-sparseness constraints on the columns of H. See help text in m-file for further information.
NMFL0_W.m: implements approximate NMF with l0-sparseness constraints on the columns of W. See help text in m-file for further information.
sparseNNLS.m: implements several functions, such as nonnegative least squares (NNLS), sparse nonnegative least squares (sNNLS) and reverse sparse nonnegative least squares (rsNNLS). See help text in m-file for further information.
experiment_SparseCoder_SyntheticData.m: reproduces experiment in section 4.1 experiment_NMFL0_H_spectrogram.m: reproduces experiment in section 4.2 experiment_NMFL0_W_ORLFaces.m: reproduces experiment in section 4.3
example_*.m: shorter application examples