Error-quantified Conformal Inference

This repository contains the data and code required to reproduce the results in ICLR2025 paper "Error-quantified Conformal Inference for Time Series", where we borrow or extend some code from PID.

Environment

Please clone this repo and run following command locally for install the environment:

conda create --name eci
pip install -r requirements.txt

Demo

Please run following command locally for local test:

cd tests
python base_test.py configs/AMZN_test.yaml
python base_plots.py results/AMZN_test.pkl

The plot results will be saved in test/plots folder. More commands can be seen in tests/expbook.ipynb. Users can modify the YAML files in the configs folder to configure the experimental settings.

Citation

If you find this work useful, you can cite it with the following BibTex entry:

@inproceedings{wu2025error,
  title={Error-quantified Conformal Inference for Time Series},
  author={Wu, Junxi and Hu, Dongjian and Bao, Yajie and Xia, Shu-tao and Zou, Changliang},
  booktitle={The Thirteenth International Conference on Learning Representations},
  year={2025}
}