Pushing the Limits of Affine Rank Minimization by Adapting Probabilistic PCA

Many applications require recovering a matrix of minimal rank within an affine constraint set, with matrix completion a notable special case. Because the problem is NP-hard in general, it is common to