JAWS-X: Addressing Efficiency Bottlenecks of Conformal Prediction Under Standard and Feedback Covariate Shift

Drew Prinster, Suchi Saria, Anqi Liu

July, 2023 Oral

@InProceedings{pmlr-v202-prinster23a,
  title = {{JAWS}-X: Addressing Efficiency Bottlenecks of Conformal Prediction Under Standard and Feedback Covariate Shift},
  author = {Prinster, Drew and Saria, Suchi and Liu, Anqi},
  booktitle = {Proceedings of the 40th International Conference on Machine Learning},
  pages = {28167--28190},
  year = {2023},
  editor = {Krause, Andreas and Brunskill, Emma and Cho, Kyunghyun and Engelhardt, Barbara and Sabato, Sivan and Scarlett, Jonathan},
  volume = {202},
  series = {Proceedings of Machine Learning Research},
  month = {23--29 Jul},
  publisher = {PMLR},
  url = {https://proceedings.mlr.press/v202/prinster23a.html}
}

Informal Summary

Accepted for an Oral Presentation at ICML 2023 (top ~2% of submissions) and building on our previous “JAWS” framework (NeurIPS 2022), this paper presents JAWS-X, a collection of methods for efficiently estimating predictive confidence intervals for black-box predictors under standard and feedback covariate shift. Our JAWS-X methods achieve distribution-free, finite-sample guarantees while flexibly balancing statistical and computational efficiency.