Abstract:This paper introduces a novel algorithm to approximate the matrix with minimum nuclear norm among all matrices obeying a set of convex constraints. This problem may be understood as the convex relaxation of a rank minimization problem and arises in many important applications as in the task of recovering a large matrix from a small subset of its entries (the famous Netflix problem). Off-the-shelf algorithms such as interior point methods are not directly amenable to large problems of this kind with over a million unknown entries. This paper develops a simple first-order and easy-to-implement algorithm that is extremely efficient at addressing problems in which the optimal solution has low rank. The algorithm is iterative, produces a sequence of matrices $\{\boldsymbol{X}^k,\boldsymbol{Y}^k\}$, and at each step mainly performs a soft-thresholding operation on the singular values of the matrix $\boldsymbol{Y}^k$. There are two remarkable features making this attractive for low-rank matrix completion problems. The first is that the soft-thresholding operation is applied to a sparse matrix; the second is that the rank of the iterates $\{\boldsymbol{X}^k\}$ is empirically nondecreasing. Both these facts allow the algorithm to make use of very minimal storage space and keep the computational cost of each iteration low. On the theoretical side, we provide a convergence analysis showing that the sequence of iterates converges. On the practical side, we provide numerical examples in which $1,000\times1,000$ matrices are recovered in less than a minute on a modest desktop computer. We also demonstrate that our approach is amenable to very large scale problems by recovering matrices of rank about 10 with nearly a billion unknowns from just about 0.4% of their sampled entries. Our methods are connected with the recent literature on linearized Bregman iterations for $\ell_1$ minimization, and we develop a framework in which one can understand these algorithms in terms of well-known Lagrange multiplier algorithms.

Finding approximately rank-one submatrices with the nuclear norm and l1 norm

Recovery of Low-Rank Matrices Based on the Rank Null Space Properties.

On the Schatten P-Quasi-norm Minimization for Low-Rank Matrix Recovery

Low Rank Matrix Recovery with Adversarial Sparse Noise

Low-Rank Approximation via Generalized Reweighted Iterative Nuclear and Frobenius Norms

A robust low-rank matrix completion based on truncated nuclear norm and Lp-norm

Nonconvex Nonsmooth Low-Rank Minimization Via Iteratively Reweighted Nuclear Norm

Fast and Accurate Matrix Completion Via Truncated Nuclear Norm Regularization

Guaranteed Minimum-Rank Solutions of Linear Matrix Equations via Nuclear Norm Minimization

Robust Recovery of Low-Rank Matrices Via Non-Convex Optimization

A Fast and Efficient Heuristic Nuclear-Norm Algorithm for Affine Rank Minimization

Matrix completion by Truncated Nuclear Norm Regularization

SVD-free Convex-Concave Approaches for Nuclear Norm Regularization.

Lower bounds on nonnegative rank via nonnegative nuclear norms

A multi-stage convex relaxation approach to noisy structured low-rank matrix recovery

On the optimal rank-1 approximation of matrices in the Chebyshev norm

Novel Alternating Update Method for Low Rank Approximation of Structured Matrices

A Singular Value Thresholding Algorithm for Matrix Completion

A Scalable, Adaptive and Sound Nonconvex Regularizer for Low-rank Matrix Learning

Fast Low-Rank Matrix Learning with Nonconvex Regularization

l1-norm regularized l1-norm best-fit lines