Trace Norm Regularization: Reformulations, Algorithms, and Multi-Task Learning
A Subspace Minimization Method for the Trust-Region Step
A Singular Value Thresholding Algorithm for Matrix Completion