Recent advances of data-driven machine learning have revolutionized fields like computer
vision, reinforcement learning, and many scientific and engineering domains. In many real …

We propose new symplectic networks (SympNets) for identifying Hamiltonian systems from
data based on a composition of linear, activation and gradient modules. In particular, we …

Abstract Reasoning about the physical world requires models that are endowed with the
right inductive biases to learn the underlying dynamics. Recent works improve …

Accurately learning the temporal behavior of dynamical systems requires models with well-
chosen learning biases. Recent innovations embed the Hamiltonian and Lagrangian …

We identify effective stochastic differential equations (SDEs) for coarse observables of fine-
grained particle-or agent-based simulations; these SDEs then provide useful coarse …

The last few years have witnessed an increased interest in incorporating physics-informed
inductive bias in deep learning frameworks. In particular, a growing volume of literature has …

Many successful methods to learn dynamical systems from data have recently been
introduced. However, ensuring that the inferred dynamics preserve known constraints, such …

Abstract Machine learning methods are widely used in the natural sciences to model and
predict physical systems from observation data. Yet, they are often used as poorly …

We propose an effective and light-weight learning algorithm, Symplectic Taylor Neural
Networks (Taylor-nets), to conduct continuous, long-term predictions of a complex …

We present a framework and algorithms to learn controlled dynamics models using neural
stochastic differential equations (SDEs)--SDEs whose drift and diffusion terms are both …