On how data are partitioned in model development and evaluation: Confronting the elephant in the room to enhance model generalization
Date
2023
Authors
Maier, H.R.
Zheng, F.
Gupta, H.
Chen, J.
Mai, J.
Savic, D.
Loritz, R.
Wu, W.
Guo, D.
Bennett, A.
Editors
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Journal article
Citation
Environmental Modelling and Software, 2023; 167:105779-1-105779-8
Statement of Responsibility
Holger R. Maier, Feifei Zheng, Hoshin Gupta, Junyi Chen, Juliane Mai, Dragan Savic, Ralf Loritz, Wenyan Wu, Danlu Guo, Andrew Bennett, Anthony Jakeman, Saman Razavi, Jianshi Zhao
Conference Name
Abstract
Models play a pivotal role in advancing our understanding of Earth’s physical nature and environmental systems, aiding in their efficient planning and management. The accuracy and reliability of these models heavily rely on data, which are generally partitioned into subsets for model development and evaluation. Surprisingly, how this partitioning is done is often not justified, even though it determines what model we end up with, how we assess its performance and what decisions we make based on the resulting model outputs. In this study, we shed light on the paramount importance of meticulously considering data partitioning in the model development and evaluation process, and its significant impact on model generalization. We identify flaws in existing data-splitting approaches and propose a forward-looking strategy to effectively confront the “elephant in the room”, leading to improved model generalization capabilities.
School/Discipline
Dissertation Note
Provenance
Description
Access Status
Rights
© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).