An implement of the NeurIPS 2021 paper: When False Positive is Intolerant: End-to-End Optimization with Low FPR for Multipartite Ranking
- Ubuntu 16.04
- CUDA 11.1
- Python 3.8.0
- Pytorch 1.7.0
See requirement.txt
.
Download datasets from Google drive and unzip it to ./data.
- Modify configs in
scripts/[dataset]/train_cba.sh
- Run the script:
sh `scripts/[dataset]/train_cba.sh`
The model and log are saved in output/[dataset]/logit_cba
by default.
- Download the pretrained model from Google drive.
- Modify configures in
scripts/[dataset]/eval_cba.yaml
: change--checkpoint
to the path where the model is saved. - Run
sh `scripts/[dataset]/eval_cba.sh`
The results might slightly differ from the above due to the environment difference in the training process.
If this code is helpful to you, please consider citing our paper:
@inproceedings{wen2021false,
title={When False Positive is Intolerant: End-to-End Optimization with Low FPR for Multipartite Ranking},
author={Wen, Peisong and Xu, Qianqian and Yang, Zhiyong and He, Yuan and Huang, Qingming},
booktitle={Thirty-Fifth Conference on Neural Information Processing Systems},
year={2021}
}