Bilevel optimization for data-driven learning of Koopman embeddings using kernel-based autoencoders

2026-10-08

AI Summary

This study addresses the reliance on manually predefined dictionaries and the lack of interpretability in neural network-based finite-dimensional approximations of Koopman operators by proposing EDMD-kDL. This method pioneers the integration of kernel methods with bilevel optimization, leveraging collocation points to learn both the kernel dictionary and Koopman embedding end-to-end from data. By replacing conventional black-box architectures with a kernel-based autoencoder, its computational complexity depends solely on the number of collocation points, ensuring excellent scalability. Experiments on sea surface temperature forecasting and video datasets demonstrate that EDMD-kDL matches or surpasses state-of-the-art models while significantly reducing dependence on training sample size, thereby achieving efficient and interpretable modeling of nonlinear dynamical systems.

Abstract

Koopman operator theory provides a linear framework for analyzing nonlinear dynamical systems and has become a major tool for data-driven modeling. A central challenge, however, is that finite-dimensional approximations computed by methods such as extended dynamic mode decomposition (EDMD) require the dictionary to be specified a priori. Recent machine-learning approaches address this limitation by learning the dictionary from data, predominantly using artificial neural network (ANN) autoencoder architectures. Although kernel methods offer an alternative with greater interpretability and tractability for theoretical analysis, they have received little attention in this setting. We introduce extended dynamic mode decomposition with kernel-based dictionary learning (EDMD-kDL), a kernel-based method for learning finite-dimensional Koopman embeddings directly from data. The method combines ideas from collocation methods and bilevel optimization to simultaneously learn a kernel dictionary and the corresponding Koopman approximation. We evaluate EDMD-kDL against state-of-the-art ANN-based approaches on a range of numerical experiments, including global sea-surface-temperature forecasting and learning directly from video data. Across all tested settings, EDMD-kDL achieves performance comparable to or better than the ANN-based methods. Moreover, in contrast to standard kernel methods, the proposed approach is scalable to large datasets by design since the size of the required kernel matrices depends on the number of collocation points rather than the size of the training dataset.

Problem

• Koopman operator
• Bilevel optimization
• Kernel methods
• Extended dynamic mode decomposition
• Dictionary learning

Innovation

• Bilevel optimization
• Koopman embeddings
• Kernel-based autoencoders
• Extended dynamic mode decomposition
• Collocation methods