PDE-NetGen 1.0: from symbolic PDE representations of physical processes to trainable neural network representations

@article{Pannekoucke2020PDENetGen1F,
  title={PDE-NetGen 1.0: from symbolic PDE representations of physical processes to trainable neural network representations},
  author={Olivier Pannekoucke and Ronan Fablet},
  journal={ArXiv},
  year={2020},
  volume={abs/2002.01029}
}
Abstract. Bridging physics and deep learning is a topical challenge. While deep learning frameworks open avenues in physical science, the design of physically-consistent deep neural network architectures is an open issue. In the spirit of physics-informed NNs, PDE-NetGen package provides new means to automatically translate physical equations, given as PDEs, into neural network architectures. PDE-NetGen combines symbolic calculus and a neural network generator. The later exploits NN-based… 

Figures from this paper

The physics informed neural networks for the unsteady Stokes problems

  • Jing YueJun Yu Li
  • Computer Science
    International Journal for Numerical Methods in Fluids
  • 2022
The paper proves the convergence of the loss function and the capability of neural networks, and combined data‐driven with model‐driven, coupled with small sample learning for solving the transient Stokes equations.

Learning Variational Data Assimilation Models and Solvers

This work develops a so‐called end‐to‐end learning approach, which explicitly relies on a variational data assimilation formulation, and shows that the variational models issued from the true Lorenz‐63 and Lorez‐96 ODE representations may not lead to the best reconstruction performance.

CDNNs: The coupled deep neural networks for coupling of the Stokes and Darcy-Forchheimer problems

An efficient deep learning method called coupled deep neural networks (CDNNs) for coupling of the Stokes and Darcy-Forchheimer problems is presented and theoretical results to guarantee the convergence of the loss function and the converge of the neural networks to the exact solution are presented.

Partial differential equations for oceanic artificial intelligence

This paper uses a model in another domain based on a data-model coupling approach to simulate and predict SST, and describes the original and modified model.

On the Complementary Role of Data Assimilation and Machine Learning: An Example Derived from Air Quality Analysis

The error statistics could benefit from a machine learning approach, while the air quality model and analysis scheme derives from physics and statistics, and out-performs the older optimum interpolation scheme using isotropic covariances with optimized covariance parameters.

Predicting porosity, permeability, and tortuosity of porous media from images by deep learning

It is shown that the CNNs can be used to predict the porosity, permeability, and tortuosity with good accuracy.

A methodology to obtain model-error covariances due to the discretization scheme from the parametric Kalman filter perspective

Abstract. This contribution addresses the characterization of the model-error covariance matrix from the new theoretical perspective provided by the parametric Kalman filter method which approximates

The Deep Learning Galerkin Method for the General Stokes Equations

Compared with traditional numerical methods, numerical results verify the theoretical analysis and show the applicability and effectiveness of the proposed DGM.

References

SHOWING 1-10 OF 25 REFERENCES

PDE-Net 2.0: Learning PDEs from Data with A Numeric-Symbolic Hybrid Deep Network

PDE-Net: Learning PDEs from Data

Numerical experiments show that the PDE-Net has the potential to uncover the hidden PDE of the observed dynamics, and predict the dynamical behavior for a relatively long time, even in a noisy environment.

Deep Hidden Physics Models: Deep Learning of Nonlinear Partial Differential Equations

  • M. Raissi
  • Computer Science
    J. Mach. Learn. Res.
  • 2018
This work puts forth a deep learning approach for discovering nonlinear partial differential equations from scattered and potentially noisy observations in space and time by approximate the unknown solution as well as the nonlinear dynamics by two deep neural networks.

Deep Neural Networks Motivated by Partial Differential Equations

A new PDE interpretation of a class of deep convolutional neural networks (CNN) that are commonly used to learn from speech, image, and video data is established and three new ResNet architectures are derived that fall into two new classes: parabolic and hyperbolic CNNs.

Residual Networks as Flows of Diffeomorphisms

This paper addresses the understanding and characterization of residual networks (ResNet), which are among the state-of-the-art deep learning architectures for a variety of supervised learning problems, and shows that this mapping component can be regarded as the numerical implementation of continuous flows of diffeomorphisms governed by ordinary differential equations.

Applications of Deep Learning to Ocean Data Inference and Subgrid Parameterization

  • T. BoltonL. Zanna
  • Environmental Science
    Journal of Advances in Modeling Earth Systems
  • 2019
The results indicate that data-driven approaches can be exploited to predict both subgrid and large-scale processes, while respecting physical principles, even when data are limited to a particular region or external forcing.

Devito (v3.1.0): an embedded domain-specific language for finite differences and geophysical exploration

Abstract. We introduce Devito, a new domain-specific language for implementing high-performance finite difference partial differential equation solvers. The motivating application is exploration

Data-driven discovery of partial differential equations

The sparse regression method is computationally efficient, robust, and demonstrated to work on a variety of canonical problems spanning a number of scientific domains including Navier-Stokes, the quantum harmonic oscillator, and the diffusion equation.

Bilinear residual Neural Network for the identification and forecasting of dynamical systems

This work considers a residual NN architecture and introduces bilinear layers to embed non-linearities which are intrinsic features of dynamical systems and demonstrates the relevance of the proposed NN-based architecture both in terms of forecasting performance and model identification.

Adam: A Method for Stochastic Optimization

This work introduces Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments, and provides a regret bound on the convergence rate that is comparable to the best known results under the online convex optimization framework.