Loss Terms and Operator Forms of Koopman Autoencoders
Machine Learning
2024-12-09 v1 Computational Physics
Abstract
Koopman autoencoders are a prevalent architecture in operator learning. But, the loss functions and the form of the operator vary significantly in the literature. This paper presents a fair and systemic study of these options. Furthermore, it introduces novel loss terms.
Cite
@article{arxiv.2412.04578,
title = {Loss Terms and Operator Forms of Koopman Autoencoders},
author = {Dustin Enyeart and Guang Lin},
journal= {arXiv preprint arXiv:2412.04578},
year = {2024}
}