Friday, 10:30 - 10:55 h, Room: H 0110


Nicolas Gillis
Fast and robust recursive algorithm for separable nonnegative matrix factorization

Coauthor: Stephen A. Vavasis


In this paper, we present an extremely fast recursive algorithm for nonnegative matrix factorization under the assumption that the nonnegative data matrix is separable (i.e., there exists a cone spanned by a small subset of the columns containing all columns). We prove that our technique is robust under any small perturbations of the data matrix, and experimentally show that it outperforms, both in terms of accuracy and speed, the state-of-the-art vertex component analysis algorithm of Nascimento and Bioucas-Dias.


Talk 1 of the invited session Fri.1.H 0110
"Recent advances in nonconvex quadratic programming with random data" [...]
Cluster 9
"Global optimization" [...]


  Payday Loans Illinois should not be obtained by people who do not have the capacity to repay the lenders. What can cause long-term use of Canadian Viagra? In the network and other sources of information, there is no reliable data on the long-term use of Viagra and its negative effects on the body.