Abstract:Data-driven approaches coupled with physical knowledge are powerful techniques to model systems. The goal of such models is to efficiently solve for the underlying field by combining measurements with known physical laws. As many systems contain unknown elements, such as missing parameters, noisy data, or incomplete physical laws, this is widely approached as an uncertainty quantification problem. The common techniques to handle all the variables typically depend on the numerical scheme used to approximate the posterior, and it is desirable to have a method which is independent of any such discretization. Information field theory (IFT) provides the tools necessary to perform statistics over fields that are not necessarily Gaussian. We extend IFT to physics-informed IFT (PIFT) by encoding the functional priors with information about the physical laws which describe the field. The posteriors derived from this PIFT remain independent of any numerical scheme and can capture multiple modes, allowing for the solution of problems which are ill-posed. We demonstrate our approach through an analytical example involving the Klein-Gordon equation. We then develop a variant of stochastic gradient Langevin dynamics to draw samples from the joint posterior over the field and model parameters. We apply our method to numerical examples with various degrees of model-form error and to inverse problems involving nonlinear differential equations. As an addendum, the method is equipped with a metric which allows the posterior to automatically quantify model-form uncertainty. Because of this, our numerical experiments show that the method remains robust to even an incorrect representation of the physics given sufficient data. We numerically demonstrate that the method correctly identifies when the physics cannot be trusted, in which case it automatically treats learning the field as a regression problem.

Consistency and convergence of simulation schemes in information field dynamics

On the Practical Applications of Information Field Dynamics

Information field dynamics for simulation scheme construction

Towards information optimal simulation of partial differential equations

Mathematical foundation of Information Field Dynamics (revised version)

Dynamical Field Inference and Supersymmetry

Iterated stochastic processes : simulation and relationship with high order partial differential equations

Monte-Carlo Simulation of a Stochastic Differential Equation

An information field theory approach to Bayesian state and parameter estimation in dynamical systems

Convergence of Dynamics on Inductive Systems of Banach Spaces

High-order conservative and accurately dissipative numerical integrators via auxiliary variables

Statistical physics of complex information dynamics

Statistical Simulation of Decaying and Forced Homogeneous Isotropic Turbulence

Hybrid Particle-Continuum Simulations Coupling Brownian Dynamics and Local Dynamic Density Functional Theory

Numerical simulations of a stochastic dynamics leading to cascades and loss of regularity: applications to fluid turbulence and generation of fractional Gaussian fields

Physics-informed Information Field Theory for Modeling Physical Systems with Uncertainty Quantification

A numerical method with properties of consistency in the energy domain for a class of dissipative nonlinear wave equations with applications to a Dirichlet boundary-value problem

Scaling up Probabilistic PDE Simulators with Structured Volumetric Information

An efficient scheme for approximating long-time dynamics of a class of non-linear models

Discrete Laplacian thermostat for flocks and swarms: the fully conservedInertial Spin Model

Numerical simulation methods and analysis for the dynamics of the time-fractional KdV equation