Guaranteed Minimum Rank Approximation from Linear Observations by Nuclear Norm Minimization with an Ellipsoidal Constraint (0903.4742v1)

Abstract: The rank minimization problem is to find the lowest-rank matrix in a given set. Nuclear norm minimization has been proposed as an convex relaxation of rank minimization. Recht, Fazel, and Parrilo have shown that nuclear norm minimization subject to an affine constraint is equivalent to rank minimization under a certain condition given in terms of the rank-restricted isometry property. However, in the presence of measurement noise, or with only approximately low rank generative model, the appropriate constraint set is an ellipsoid rather than an affine space. There exist polynomial-time algorithms to solve the nuclear norm minimization with an ellipsoidal constraint, but no performance guarantee has been shown for these algorithms. In this paper, we derive such an explicit performance guarantee, bounding the error in the approximate solution provided by nuclear norm minimization with an ellipsoidal constraint.