Code Repository

We have set up a consolidated code repository where we will share updated implementations of various algorithms we proposed. The goal is to make them easy to use and expand, both by researchers and practitoners.

The repository is found at

https://github.com/MLDS-NUS

Currently, we have a repository released on Koopman decomposition using deep learning, following our original work of Li, Q., Dietrich, F., Bollt, E. M., & Kevrekidis, I. G. (2017). Extended dynamic mode decomposition with dictionary learning: A data-driven adaptive spectral decomposition of the Koopman operator. Chaos: An Interdisciplinary Journal of Nonlinear Science, 27(10), 103111..

More implementations will be added in the future!

News Article at SIAM News

In our SIAM news article on Machine Learning and Dynamical Systems, we introduce some of our recent work on the approximation theory and related aspects, on the intersection of machine learning and dynamical systems. Note that this field is rapidly growing and due to length limitations this article is by no means a review of the progress in the field, but rather as a short exposition of our recent work in this area.

For more comprehensive coverage of the key contributions to this effort, interested readers are referred to this interest group hosted at the Alan Turing Institute, organized by Profs Robert MacKay and Boumediene Hamzi.