Holographic Factorization Machines

An implementation for Holographic Factorization Machines (HFM).

As an example, trained with data 'rating.csv' of Anime Recommendations Database downloading from kaggle.

Data format:

user_id, anime_id, rating

Only used feature engineering processes as:

1. Turn user_id from 1~M into 0~M-1
2. Turn anime_id from 1~N into 0~N-1
3. Turn rating from 1~10 into 0.1~1.0 by divided by 10.