Coarse-grained and emergent distributed parameter systems from data

TL;DR

This paper presents a data-driven approach combining manifold learning and neural networks to identify PDEs and their variables from spatiotemporal data without prior knowledge of the variables.

Contribution

It introduces a novel method using Diffusion Maps and neural networks to discover PDE structures and variables directly from data, even when they are unknown beforehand.

Findings

Successfully identified PDEs from simulated data

Demonstrated variable discovery using local particle distributions

Connected data-driven PDE discovery with multiscale computation tools

Abstract

We explore the derivation of distributed parameter system evolution laws (and in particular, partial differential operators and associated partial differential equations, PDEs) from spatiotemporal data. This is, of course, a classical identification problem; our focus here is on the use of manifold learning techniques (and, in particular, variations of Diffusion Maps) in conjunction with neural network learning algorithms that allow us to attempt this task when the dependent variables, and even the independent variables of the PDE are not known a priori and must be themselves derived from the data. The similarity measure used in Diffusion Maps for dependent coarse variable detection involves distances between local particle distribution observations; for independent variable detection we use distances between local short-time dynamics. We demonstrate each approach through an illustrative established PDE example. Such variable-free, emergent space identification algorithms connect naturally with equation-free multiscale computation tools.