inhenn / iheartsteam

Recommender system on steam data

iHeartsteam

Recommender system on steam data

data we use: steam data

You may want to start from preprocessing data, which can be done use preprocessing.py in the same repo.

In this project, we implement two models for the steam game recommender system: Factorization Machine and ANN with LSH Both two models runs well, yet compared with the method implemented in the package, we still need to improve efficiency and code style. Further more, we would like to emsemble both method for testing and improving coverage, which we lack of time to evaluate.

Please refer to final report.ipynb for our report.