- "Data poisoning Attack on Factorization-Based Collaborative Filtering" NIPS 2016
- Github respository : "recommend"
- Github respository : "movielens"
- ALS_optimize.py: compute optimize solution using malicious data and normal data.
- ALS_optimize_old.py: compute optimize solution only using normal data.
- compute_grad.py: compute two parts of gradient using the formulations in the paper.
- dataset.py: functions of load dataset and build user item matrix.
- evaluation.py: predict data and compure RMSE
- main.py: the implement of PGA