This repository is a recommender system tutorial on a class of collaborative filtering models called low-rank approximation (LRA), also known as matrix factorization (MF). In the tutorial you will be introduced to the basic concepts, implement the model and apply it to the MovieLens 1M dataset, containing one million ratings given by about 6k users to 4k movies.

To get started, clone or download this repository, start a Terminal (macOS) or Command Prompt (Windows) and create the tutorial's conda environment with conda env create --force. Then activate it with source activate recsys-env (macOS) or activate recsys-env (Windows), and start a Jupyter notebook server with jupyter notebook.