Copula-based conformal prediction for Multi-Target Regression
Machine Learning
2021-01-29 v1 Artificial Intelligence
Machine Learning
Abstract
There are relatively few works dealing with conformal prediction for multi-task learning issues, and this is particularly true for multi-target regression. This paper focuses on the problem of providing valid (i.e., frequency calibrated) multi-variate predictions. To do so, we propose to use copula functions applied to deep neural networks for inductive conformal prediction. We show that the proposed method ensures efficiency and validity for multi-target regression problems on various data sets.
Cite
@article{arxiv.2101.12002,
title = {Copula-based conformal prediction for Multi-Target Regression},
author = {Soundouss Messoudi and Sébastien Destercke and Sylvain Rousseau},
journal= {arXiv preprint arXiv:2101.12002},
year = {2021}
}
Comments
17 pages, 8 figures, under review