Two-dimensional semi-nonnegative matrix factorization for clustering

In this paper, we propose a new Semi-Nonnegative Matrix Factorization method for 2-dimensional (2D) data, named TS-NMF. It overcomes the drawback of existing methods that seriously damage spatial information data by converting 2D to vectors in preprocessing step. particular, projection matrices are sought under guidance building representations, such is retained and projections enhanced goal clustering, which helps construct optimal directions. Moreover, exploit nonlinear structures manifold constructed projected subspace, adaptively updated according less afflicted with noise outliers thus more representative space. Hence, seeking projections, learning seamlessly integrated single model, mutually enhance other lead powerful representation. Comprehensive experimental results verify effectiveness TS-NMF comparison several state-of-the-art algorithms, suggests high potential proposed real world applications.