GP-lite is a lightweight, educational toolkit for Gaussian Processes in Python. It focuses on clarity, minimalism, and ease of modification. Its intended use is teaching, demonstrations, courseworks, and understanding the maths and basic concepts behind GPs in general.
The corresponding Python package is gplite.
GP-lite was originally developed for the NeurIPS 2021 tutorial "The Art of Gaussian Processes: Classic and Contemporary" by César L. Mattos and Felipe Tobar. This repository is a continuous update of the toolbox.
Author: Felipe Tobar