CAS Seminar

Sergiy Vasylkevich

(Jacobs University)

"Non-negative Matrix factorization with factorizable feature matrix"

Date: Tue, October 30, 2012
Time: 11:15
Place: Research I Seminar Room

Abstract: Non-negative Matrix factorization (NFM) algorithms are widely used in image recognition, collaborative filtering, and environmental monitoring applications. NMF represents data as a linear combination of basis features, which are identified by the algorithm. I give an introduction to NMF theory and present our extension of NMF and weighted NMF to the problems where feature matrix must be factorizable with a prescribed co-factor.