New Python Package: GPyConform – Gaussian Process Regression (Full) CP

[h.papad...@frederick.ac.cy]

Dear all,

I am thrilled to announce the release of GPyConform, a Python package that extends the GPyTorch library by implementing (Full) Conformal Prediction for Gaussian Process Regression. Designed to work seamlessly with Exact Gaussian Process (GP) models, GPyConform enhances GPyTorch by introducing the capability to generate and evaluate both symmetric and asymmetric Conformal Prediction Intervals.

GPyConform is open-source and available for download and download and installation via PyPI and conda-forge. You can access the package and its documentation here:

GitHub Repository: https://github.com/harrisp/GPyConform
Documentation: https://gpyconform.readthedocs.io/en/latest/

I look forward to any feedback, should you or some of your colleagues/students would be interested in trying it out.

Warm regards,

Harris Papadopoulos, Associate Professor

Dep. of Electrical Engineering, Computer Engineering & Informatics,
School of Engineering, Frederick University, Nicosia, Cyprus

Head, Computational Intelligence (COIN) Research Lab