Fast and Accurate Matrix Completion via Truncated Nuclear Norm Regularization
Fast and Accurate Matrix Completion via Truncated Nuclear Norm Regularization is a scholarly work by Xuelong Li, published in 2013 in ''IEEE Transactions on Pattern Analysis and Machine Intelligence''. The main subjects of the publication include singular value, rate of convergence, line search, optimization problem, rank, image segmentation, low-rank approximation, matrix completion, matrix, computer science, mathematical optimization, inpainting, compressed sensing, norm, mathematics, facial recognition system, matrix norm, and algorithm. The authors propose to achieve a better approximation to the rank of matrix by truncated nuclear norm, which is given by the nuclear norm subtracted by the sum of the largest few singular values.