A Fast Parallel Stochastic Gradient Method for Matrix Factorization in Shared Memory Systems

Matrix factorization is known to be an effective method for recommender systems that are given only the ratings from users to items. Currently, stochastic gradient (SG) method is one of the most po...