Recommender systems are widely used by web retailers to suggest products to their customers; for example, Amazon tells you what other customers who purchased this item were interested in or Netflix suggests TV serials and movies based on what you have watched and what other Netflix users with the same interest have watched. These recommender systems work on the basis of collaborative filtering. In collaborative filtering, the system builds a model from a user's past behavior. We will use the RBM, made in the previous recipe, to build a recommender system using collaborative filtering to recommend movies. An important challenge in this work is that most users will not rate all products/movies, thus most data is missing. If there are M products and N users, then we need to build an array, N x M, which contains the known ratings of the users and makes all unknown values zero.