Ensemble Regularized Adaptive Prediction Set ERAPS

Implementation and experiments based on the paper Conformal prediction set for time-series. The current paper is strongly accepted by the Workshop on Distribution-Free Uncertainty Quantification at ICML 2022

Please direct all implementation-related inquiries to Chen Xu @ cxu310@gatech.edu.

Citation:

  @misc{xuERPAS2022,
  doi = {10.48550/ARXIV.2206.07851},
  url = {https://arxiv.org/abs/2206.07851},
  author = {Xu, Chen and Xie, Yao},
  keywords = {Machine Learning (stat.ML), Machine Learning (cs.LG), Methodology (stat.ME), FOS: Computer and information sciences, FOS: Computer and information sciences},
  title = {Conformal prediction set for time-series},
  publisher = {arXiv},
  year = {2022},
  copyright = {Creative Commons Attribution 4.0 International}
}

Documentation

• Executing the file ERAPS_test.py reproduces our results.
• The Table below compares ERAPS vs. three competing methods on the MelbournePedestrian dataset, where we can see that ERAPS always maintains valid marginala coverage with much smaller prediction sets in some cases.

Comparison of ERAPS vs. competitors