Causal hybrid modeling with double machine learning ...
Hybrid modeling integrates machine learning with scientific knowledge to enhance interpretability, generalization, and adherence to natural laws. Nevertheless, equifinality and regularization biases pose challenges in hybrid modeling to achieve these purposes. This paper introduces a novel approach...
Main Authors: | , , , , |
---|---|
Format: | Article in Journal/Newspaper |
Language: | unknown |
Published: |
arXiv
2024
|
Subjects: | |
Online Access: | https://dx.doi.org/10.48550/arxiv.2402.13332 https://arxiv.org/abs/2402.13332 |
id |
ftdatacite:10.48550/arxiv.2402.13332 |
---|---|
record_format |
openpolar |
spelling |
ftdatacite:10.48550/arxiv.2402.13332 2024-06-09T07:45:36+00:00 Causal hybrid modeling with double machine learning ... Cohrs, Kai-Hendrik Varando, Gherardo Carvalhais, Nuno Reichstein, Markus Camps-Valls, Gustau 2024 https://dx.doi.org/10.48550/arxiv.2402.13332 https://arxiv.org/abs/2402.13332 unknown arXiv arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ Machine Learning cs.LG Methodology stat.ME FOS Computer and information sciences CreativeWork article Preprint Article 2024 ftdatacite https://doi.org/10.48550/arxiv.2402.13332 2024-05-13T10:48:18Z Hybrid modeling integrates machine learning with scientific knowledge to enhance interpretability, generalization, and adherence to natural laws. Nevertheless, equifinality and regularization biases pose challenges in hybrid modeling to achieve these purposes. This paper introduces a novel approach to estimating hybrid models via a causal inference framework, specifically employing Double Machine Learning (DML) to estimate causal effects. We showcase its use for the Earth sciences on two problems related to carbon dioxide fluxes. In the $Q_{10}$ model, we demonstrate that DML-based hybrid modeling is superior in estimating causal parameters over end-to-end deep neural network (DNN) approaches, proving efficiency, robustness to bias from regularization methods, and circumventing equifinality. Our approach, applied to carbon flux partitioning, exhibits flexibility in accommodating heterogeneous causal effects. The study emphasizes the necessity of explicitly defining causal graphs and relationships, advocating ... Article in Journal/Newspaper DML DataCite Metadata Store (German National Library of Science and Technology) |
institution |
Open Polar |
collection |
DataCite Metadata Store (German National Library of Science and Technology) |
op_collection_id |
ftdatacite |
language |
unknown |
topic |
Machine Learning cs.LG Methodology stat.ME FOS Computer and information sciences |
spellingShingle |
Machine Learning cs.LG Methodology stat.ME FOS Computer and information sciences Cohrs, Kai-Hendrik Varando, Gherardo Carvalhais, Nuno Reichstein, Markus Camps-Valls, Gustau Causal hybrid modeling with double machine learning ... |
topic_facet |
Machine Learning cs.LG Methodology stat.ME FOS Computer and information sciences |
description |
Hybrid modeling integrates machine learning with scientific knowledge to enhance interpretability, generalization, and adherence to natural laws. Nevertheless, equifinality and regularization biases pose challenges in hybrid modeling to achieve these purposes. This paper introduces a novel approach to estimating hybrid models via a causal inference framework, specifically employing Double Machine Learning (DML) to estimate causal effects. We showcase its use for the Earth sciences on two problems related to carbon dioxide fluxes. In the $Q_{10}$ model, we demonstrate that DML-based hybrid modeling is superior in estimating causal parameters over end-to-end deep neural network (DNN) approaches, proving efficiency, robustness to bias from regularization methods, and circumventing equifinality. Our approach, applied to carbon flux partitioning, exhibits flexibility in accommodating heterogeneous causal effects. The study emphasizes the necessity of explicitly defining causal graphs and relationships, advocating ... |
format |
Article in Journal/Newspaper |
author |
Cohrs, Kai-Hendrik Varando, Gherardo Carvalhais, Nuno Reichstein, Markus Camps-Valls, Gustau |
author_facet |
Cohrs, Kai-Hendrik Varando, Gherardo Carvalhais, Nuno Reichstein, Markus Camps-Valls, Gustau |
author_sort |
Cohrs, Kai-Hendrik |
title |
Causal hybrid modeling with double machine learning ... |
title_short |
Causal hybrid modeling with double machine learning ... |
title_full |
Causal hybrid modeling with double machine learning ... |
title_fullStr |
Causal hybrid modeling with double machine learning ... |
title_full_unstemmed |
Causal hybrid modeling with double machine learning ... |
title_sort |
causal hybrid modeling with double machine learning ... |
publisher |
arXiv |
publishDate |
2024 |
url |
https://dx.doi.org/10.48550/arxiv.2402.13332 https://arxiv.org/abs/2402.13332 |
genre |
DML |
genre_facet |
DML |
op_rights |
arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ |
op_doi |
https://doi.org/10.48550/arxiv.2402.13332 |
_version_ |
1801375038530650112 |