Codes and data to reproduce the tables and figures of
"Negative impact of heavy-tailed uncertainty and error distributions on the reliability of calibration statistics for machine learning regression tasks"
a revised and expanded version of
"How to validate average calibration for machine learning regression tasks ?" by P. Pernot (https://arxiv.org/abs/2402.10043v2)
Codes in "R" repertory
- "0_main.R" the main code, calls several scripts.
Note: The dataset JAC2024 that is referred to in JAC2024 is too big to be stored here.