Low-rank matrix factorization finds applications in large number of problems in signal processing and machine learning. Stochastic gradient descent (SGD) is a standard technique for solving large scale matrix factorization ...