We solve separable NMF problem under convex relaxation in the form of self-dictionary multiple measurement vectors (SD-MMV), using Frank-Wolfe algorithm. It possesses desirable identifiablity properties even under noisy case while only requires a linear growth to data size in memory.
ductri/Frank-Wolfe-based-method-for-SD-MMV
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.