Generalizing to New Physical Systems via Context-Informed Dynamics Model

Matthieu Kirchmeyer, Yuan Yin, Jeremie Dona, Nicolas Baskiotis, Alain Rakotomamonjy, Patrick Gallinari
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:11283-11301, 2022.

Abstract

Data-driven approaches to modeling physical systems fail to generalize to unseen systems that share the same general dynamics with the learning domain, but correspond to different physical contexts. We propose a new framework for this key problem, context-informed dynamics adaptation (CoDA), which takes into account the distributional shift across systems for fast and efficient adaptation to new dynamics. CoDA leverages multiple environments, each associated to a different dynamic, and learns to condition the dynamics model on contextual parameters, specific to each environment. The conditioning is performed via a hypernetwork, learned jointly with a context vector from observed data. The proposed formulation constrains the search hypothesis space for fast adaptation and better generalization across environments with few samples. We theoretically motivate our approach and show state-of-the-art generalization results on a set of nonlinear dynamics, representative of a variety of application domains. We also show, on these systems, that new system parameters can be inferred from context vectors with minimal supervision.

Cite this Paper

BibTeX
@InProceedings{pmlr-v162-kirchmeyer22a, title = {Generalizing to New Physical Systems via Context-Informed Dynamics Model}, author = {Kirchmeyer, Matthieu and Yin, Yuan and Dona, Jeremie and Baskiotis, Nicolas and Rakotomamonjy, Alain and Gallinari, Patrick}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {11283--11301}, year = {2022}, editor = {Chaudhuri, Kamalika and Jegelka, Stefanie and Song, Le and Szepesvari, Csaba and Niu, Gang and Sabato, Sivan}, volume = {162}, series = {Proceedings of Machine Learning Research}, month = {17--23 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v162/kirchmeyer22a/kirchmeyer22a.pdf}, url = {https://proceedings.mlr.press/v162/kirchmeyer22a.html}, abstract = {Data-driven approaches to modeling physical systems fail to generalize to unseen systems that share the same general dynamics with the learning domain, but correspond to different physical contexts. We propose a new framework for this key problem, context-informed dynamics adaptation (CoDA), which takes into account the distributional shift across systems for fast and efficient adaptation to new dynamics. CoDA leverages multiple environments, each associated to a different dynamic, and learns to condition the dynamics model on contextual parameters, specific to each environment. The conditioning is performed via a hypernetwork, learned jointly with a context vector from observed data. The proposed formulation constrains the search hypothesis space for fast adaptation and better generalization across environments with few samples. We theoretically motivate our approach and show state-of-the-art generalization results on a set of nonlinear dynamics, representative of a variety of application domains. We also show, on these systems, that new system parameters can be inferred from context vectors with minimal supervision.} }
Endnote
%0 Conference Paper %T Generalizing to New Physical Systems via Context-Informed Dynamics Model %A Matthieu Kirchmeyer %A Yuan Yin %A Jeremie Dona %A Nicolas Baskiotis %A Alain Rakotomamonjy %A Patrick Gallinari %B Proceedings of the 39th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2022 %E Kamalika Chaudhuri %E Stefanie Jegelka %E Le Song %E Csaba Szepesvari %E Gang Niu %E Sivan Sabato %F pmlr-v162-kirchmeyer22a %I PMLR %P 11283--11301 %U https://proceedings.mlr.press/v162/kirchmeyer22a.html %V 162 %X Data-driven approaches to modeling physical systems fail to generalize to unseen systems that share the same general dynamics with the learning domain, but correspond to different physical contexts. We propose a new framework for this key problem, context-informed dynamics adaptation (CoDA), which takes into account the distributional shift across systems for fast and efficient adaptation to new dynamics. CoDA leverages multiple environments, each associated to a different dynamic, and learns to condition the dynamics model on contextual parameters, specific to each environment. The conditioning is performed via a hypernetwork, learned jointly with a context vector from observed data. The proposed formulation constrains the search hypothesis space for fast adaptation and better generalization across environments with few samples. We theoretically motivate our approach and show state-of-the-art generalization results on a set of nonlinear dynamics, representative of a variety of application domains. We also show, on these systems, that new system parameters can be inferred from context vectors with minimal supervision.
APA
Kirchmeyer, M., Yin, Y., Dona, J., Baskiotis, N., Rakotomamonjy, A. & Gallinari, P.. (2022). Generalizing to New Physical Systems via Context-Informed Dynamics Model. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:11283-11301 Available from https://proceedings.mlr.press/v162/kirchmeyer22a.html.

Related Material