A multilayer approach to nonnegative matrix factorisation algorithms is proposed. It considerably improves their performance; especially if a problem is ill-conditioned, or data are badly scaled, and projected gradient algorithms are used. This is fully confirmed by extensive simulations with diverse types of data in application to blind source separation.
No takes yet. Share an insight, caveat, or question.
Cichocki et al. (2006) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: