Temporally-Consistent Koopman Autoencoders for Forecasting Dynamical Systems

arXiv:2403.12335v1 Announce Type: new
Abstract: Absence of sufficiently high-quality data often poses a key challenge in data-driven modeling of high-dimensional spatio-temporal dynamical systems. Koopman Autoencoders (KAEs) harness the expressivity of deep neural networks (DNNs), the dimension reduction capabilities of autoencoders, and the spectral properties of the Koopman operator to learn a reduced-order feature space with simpler, linear dynamics. However, the effectiveness of KAEs is hindered by limited and noisy training datasets, leading to poor generalizability. To address this, we introduce …

abstract arxiv autoencoders capabilities challenge consistent cs.lg data data-driven feature forecasting harness key learn modeling networks neural networks quality quality data systems temporal type