pouyapez / grr-matrix-factorization Goto Github PK

To train Movielens dataset, download the training and test data from:

https://grouplens.org/datasets/movielens/

and put the data in the "data" folder. Then assign your low rank approximation value (k) and link function (link_func) in "movie_lens.py" and run it. After training to evaluate the results, choose the related link function and proper directory (where you saved your model) in "movie_lens_test.py" and run it to calculate RMSE and percentage accuracy on the test data.