Tuesday, 13:15 - 13:40 h, Room: H 3004


Stephen Vavasis
Identifying k large submatrices using convex programming

Coauthors: Venkat Chandrasekaran, Xuan Vinh Doan


We consider the problem of identifying k large approximately rank-one submatrices of a nonnegative data matrix. Stated in a certain manner, this problem is NP-hard, but has important applications in data mining. In particular it is a version of the well-known nonnegative matrix factorization, which has been applied to document classification, image decomposition, and analysis of biochemical experiments. We prove that if the data is constructed according to a certain randomized model, then the k blocks can be recovered in polynomial time via convex relaxation.


Talk 1 of the invited session Tue.2.H 3004
"Cone factorizations and lifts of convex sets" [...]
Cluster 2
"Combinatorial optimization" [...]


  Payday Loans Wisconsin. Therefore, we can say that the active substances in its composition are more perfectly mixed. Vardenafil is not only present in the original Order Levitra, but also as part of its analogs.