Python Library PyCalib

Authors

Affiliations

Miquel Perello Nieto

University of Bristol

Telmo Silva Filho

Federal University of Paraiba

Hao Song

University of Bristol

We have developed a Python library that offers multiple tools to assess probabilistic classifiers in terms of calibration, and provides useful function to calibrate models that follow the Scikit-learn BaseEstimator standar.

The documentation for the library can be found at classifier-calibration.github.io/PyCalib/ and the library can be installed from the PyPi package repository.

Citation

If you use the library for a research publication you may want to cite the publication below, which provides a survey of classifier calibration and mentions this Python library.

@Article{SilvaFilho2023,
author={Silva Filho, Telmo
and Song, Hao
and Perello-Nieto, Miquel
and Santos-Rodriguez, Raul
and Kull, Meelis
and Flach, Peter},
title={Classifier calibration: a survey on how to assess and improve predicted class probabilities},
journal={Machine Learning},
year={2023},
month={May},
day={16},
issn={1573-0565},
doi={10.1007/s10994-023-06336-7},
url={https://doi.org/10.1007/s10994-023-06336-7}
}