Literature Survey: Non-Negative Matrix Factorization Export

@misc{citeulike:3140579, abstract = {Abstract. This article surveys recent research on Non-Negative Matrix Factorization (NNMF), a relatively new technique for dimensionality reduction. It is based on the idea that, in many data-processing tasks, negative numbers are physically meaningless. The NNMF technique addresses this problem by placing non-negativity constraints on the data model. I discuss the applications of NNMF, the proposed algorithms and the qualitative results. Since many of the algorithms suggested for NNMF seem to lack a firm theoret-ical foundation, this article also surveys techniques for proving that iterative algorithms converge. It concludes with a description of additional investigations which are presently underway. 2. NON-NEGATIVE MATRIX FACTORIZATION 3 1.}, citeulike-article-id = {3140579}, citeulike-linkout-0 = {http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.84.9645}, keywords = {learning, machine, multivariate, nmf, nnma}, posted-at = {2008-08-20 15:22:17}, priority = {2}, title = {Literature Survey: Non-Negative Matrix Factorization}, url = {http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.84.9645} }

TY - ELEC ID - citeulike:3140579 L3 - citeulike-article-id:3140579 N2 - Abstract. This article surveys recent research on Non-Negative Matrix Factorization (NNMF), a relatively new technique for dimensionality reduction. It is based on the idea that, in many data-processing tasks, negative numbers are physically meaningless. The NNMF technique addresses this problem by placing non-negativity constraints on the data model. I discuss the applications of NNMF, the proposed algorithms and the qualitative results. Since many of the algorithms suggested for NNMF seem to lack a firm theoret-ical foundation, this article also surveys techniques for proving that iterative algorithms converge. It concludes with a description of additional investigations which are presently underway. 2. NON-NEGATIVE MATRIX FACTORIZATION 3 1. TI - Literature Survey: Non-Negative Matrix Factorization KW - learning KW - machine KW - multivariate KW - nmf KW - nnma UR - http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.84.9645 ER -