#Recommender-By-Matrix-Factorization This is a Movie Recommender System Based on Matrix Factorization, train and test on the movielens100k.mat dataset. The dataset for this project comes from MovieLens (http://grouplens.org/). It consists of 100,000 ratings from 1000 users on 1700 movies. The rating values go from 0 to 5. The dataset is split into train and test data, around 95,000 and 5,000 ratings respectively.
With all the parameters well set, a RMSE of around 0.8008 (actually MSE according to the formula provided) can be achieved.