Using the Mean Absolute Percentage Error for Regression Models
Machine Learning
2015-06-16 v1 Machine Learning
Abstract
We study in this paper the consequences of using the Mean Absolute Percentage Error (MAPE) as a measure of quality for regression models. We show that finding the best model under the MAPE is equivalent to doing weighted Mean Absolute Error (MAE) regression. We show that universal consistency of Empirical Risk Minimization remains possible using the MAPE instead of the MAE.
Cite
@article{arxiv.1506.04176,
title = {Using the Mean Absolute Percentage Error for Regression Models},
author = {Arnaud De Myttenaere and Boris Golden and Bénédicte Le Grand and Fabrice Rossi},
journal= {arXiv preprint arXiv:1506.04176},
year = {2015}
}
Comments
European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN), Apr 2015, Bruges, Belgium. 2015, Proceedings of the 23-th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2015)