Fast smooth rank function approximation based on matrix tri-factorization

摘要: Abstract Recently, Smooth Rank Function (SRF) is proposed for matrix completion problem. The main idea of this algorithm based on a continuous and differentiable approximation the rank function. However, it need to deal with singular value decomposition in each iteration, which consumes much time large matrix. In paper, by utilizing tri-factorization matrix, fast method SRF proposed. Then, our method, adaptive smooth function presented appropriate estimation. We mathematically prove convergence method. Experimental results show that improves running significantly. Furthermore, outperforms other existing approaches most cases.