Co-clustering under Nonnegative Matrix Tri-Factorization

摘要

The nonnegative matrix tri-factorization (NMTF) approach has recently been shown to be useful and effective to tackle the co-clustering. In this work, we embed this problem in the NMF framework and we derive from the double k-means objective function a new formulation of the criterion. To optimize it, we develop two algorithms based on two multiplicative update rules. In addition we show that the double k-means is equivalent to algebraic problem of NMF under some suitable constraints. Numerical experiments on simulated and real datasets demonstrate the interest of our approach.