I've used the lightFM library to train a hybrid content-based + collaborative algorithm that uses the WARP loss function on the movielens dataset. Movielens is a great source of movies and ratings from thousands of users. The model will create personalized movie recommendations in a setup environment (Terminal)
Bufkin/Movie_Recommendation_System
I've used the LightFM library to train a hybrid content-based + collaborative algorithm that uses the WARP loss function on the movielens dataset. Movielens is a great source of movies and ratings from thousands of users. The model will create personalized movie recommendations in a setup environment (Terminal)
Python