This is repo of paper "CoreNet: Leveraging Context-Aware Representations via MLP Networks for CTR Prediction".
Datasets are downloaded from:
- Criteo, Frappe, MovieLens 2M: https://github.com/WeiyuCheng/AFN-AAAI-20
- Avazu: https://www.kaggle.com/c/avazu-ctr-prediction
The code has been tested running under Python 3.8. The required packaged for reproducing environment could be found in requirements.txt
pip install -r requirements.txt
run
python src/preprocess/run_preprocess.py --dataset {name_dataset}
run
python src/train.py --dataset {name_dataset}