Problems with non-negativity constrains have recently attracted a great deal of interest. Non-negativity constraints arise naturally in many applications, and are often necessary for proper interpretation. Furthermore, these constrains provide an intrinsic sparsity that may be of value in certain situations. Two common problems that have gathered notable attention are the non-negative least squares (NNLS) problem, and the nonnegative matrix factorization (NMF) problem. In this paper, a method to solve the NNLS problem in an adaptive way is discussed. Additionally, possible ways to apply this, and other related method, to adaptive NMF problems are discussed.
University of Minnesota M.S.E.E. thesis. June 2014. Major: Electrical Engineering. Advisor: Nikolaos Sidiropoulos. 1 computer file (PDF); v, 38 pages.
Adaptive Non-negative Least Squares with Applications to Non-Negative Matrix Factorization.
Retrieved from the University of Minnesota Digital Conservancy,
Content distributed via the University of Minnesota's Digital Conservancy may be subject to additional license and use restrictions applied by the depositor.