Research Article
Relationship Matrix Nonnegative Decomposition for Clustering
Algorithm 1
Relationship matrix nonnegative decomposition (RMND).
Input: Positive matrix , and a positive integer . | Output: Nonnegative factor matrices , and . | Learning Procedure | (S1) Initialize , and to random positive matrices, and normalize each column of to one. | (S2) Repeat the iterations until convergence: | (1) . | (2) . | (3) and where is a diagonal matrix with . | (4) Repeatedly select small positive constant until the objective function is decreased. | (i) . | (ii) Project each column of to be nonnegative vector with unit norm. | (5) . | Above, and denote elementwise multiplication and division, respectively. |
|