aloysiogl / conforme

Repository for the paper "ConForME: Multi-horizon conformal time series forecasting"

Paper: ConForME: Multi-horizon conformal time series forecasting

Table of contents:

• Abstract
• Installation
• Download data
• Usage
• Code structure
• License

TLDR

You'll need poetry and git-lfs to run and download the data. Then, run the following commands from the root of the repo to install, pull the data, and run the experiments:

poetry install
git lfs install && git lfs pull
poetry run python scripts/run_experiments_tables.py
poetry run python scripts/run_experiments_betas.py

JSON files with the results will be saved in the results directory. Further instructions detailing how to make these results more readable can be found in here.

How to cite

@inproceedings{conforme2024,
  author       = {Aloysio Galvao Lopes and
                  Eric Goubault and
                  Laurent Pautet and
                  Sylvie Putot},
  editor       = {Simone Vantini and
                  Matteo Fontana and
                  Aldo Solari and
                  Henrik Bostr{\"{o}}m and
                  Lars Carlsson},
  title        = {ConForME: Multi-horizon conformal time series forecasting},
  booktitle    = {Conformal and Probabilistic Prediction with Applications, 9-11 September
                  2024, Milan, Italy},
  series       = {Proceedings of Machine Learning Research},
  volume       = {230},
  publisher    = {{PMLR}},
  year         = {2024},
}