From Linear combinationTo Recommendation algorithm

Introduce a use case of a collaborative (based) filtering based recommendation system via deep learning. This article will not use the mathematical terms of linear algebra or term knowledge, which are occasionally mentioned, and do not count these too much. A question is raised here, can we use a kind of supervised learning algorithm to predict the rating a user might give by a user-ID and a movie-ID, or a user-ID and a lecture-ID? We then compare the estimated rating or quantity of interest with a reasonable threshold to decide whether we should recommend the item to the user or not. Recall that the vector (in row-wise) X0 and X1 can be stretched by Beta vector (two factors) on their way vertically and horizontally.