Converged Algorithms for Orthogonal Nonnegative Matrix Factorizations

About

This paper proposes uni-orthogonal and bi-orthogonal nonnegative matrix factorization algorithms with robust convergence proofs. We design the algorithms based on the work of Lee and Seung [1], and derive the converged versions by utilizing ideas from the work of Lin [2]. The experimental results confirm the theoretical guarantees of the convergences.

Andri Mirzal · 2010

Related benchmarks

Task	Dataset	Result
Clustering	Reuters2	Average Runtime38.367	7
Clustering	Reuters4	Average Runtime86.745	7
Clustering	Reuters6	Average Running Time75.432	7
Clustering	Reuters8	Average Runtime56.464	7
Clustering	Reuters10	Average Running Time601.57	7
Document Clustering	Reuters2	F-measure84.823	7
Document Clustering	Reuters4	Fmeasure57.989	7
Document Clustering	Reuters6	F-measure48.444	7
Document Clustering	Reuters8	Fmeasure42.996	7
Document Clustering	Reuters2	Average Entropy0.94498	7