Recommender system based on embeddings for movielens dataset

Implemented deep learning models create user and movie embeddings. Using neural network approaches, the models build vector representations of users and movies (embeddings).The resulting embeddings can be used to determine the similarity of movies and users.

How to run

1. Download MovieLens dataset and put it in "ml-latest/" folder
2. Run preprocessing.ipynb
3. Run model from models.ipynb or transformer.ipynb