Unsupervised Learning of Solutions to Differential Equations with Generative Adversarial Networks

Abstract

Solutions to differential equations are of significant scientific and engineering relevance. Recently, there has been a growing interest in solving differential equations with neural networks. This work develops a novel method for solving differential equations with unsupervised neural networks that applies Generative Adversarial Networks (GANs) to \emph{learn the loss function} for optimizing the neural network. We present empirical results showing that our method, which we call Differential Equation GAN (DEQGAN), can obtain multiple orders of magnitude lower mean squared errors than an alternative unsupervised neural network method based on (squared) $L_2$, $L_1$, and Huber loss functions. Moreover, we show that DEQGAN achieves solution accuracy that is competitive with traditional numerical methods. Finally, we analyze the stability of our approach and find it to be sensitive to the selection of hyperparameters, which we provide in the appendix. Code available at https://github.com/dylanrandle/denn. Please address any electronic correspondence to [email protected].

References in corpus (17)

• Adam: A Method for Stochastic Optimization
• Conditional Generative Adversarial Nets
• Playing Atari with Deep Reinforcement Learning
• Deep Residual Learning for Image Recognition
• DGM: A deep learning algorithm for solving partial differential equations
• Solving high-dimensional partial differential equations using deep learning
• Improved Training of Wasserstein GANs
• BEGAN: Boundary Equilibrium Generative Adversarial Networks
• Autoencoding beyond pixels using a learned similarity metric
• Artificial Neural Network Methods in Quantum Mechanics
• Distributional Reinforcement Learning with Quantile Regression
• Solving differential equations with neural networks: Applications to the calculation of cosmological phase transitions
• Physics-Informed Generative Adversarial Networks for Stochastic Differential Equations
• Physical Symmetries Embedded in Neural Networks
• FiniteNet: A Fully Convolutional LSTM Network Architecture for Time-Dependent Partial Differential Equations
• Solving differential equations with unknown constitutive relations as recurrent neural networks
• Neural Network Methods for Boundary Value Problems Defined in Arbitrarily Shaped Domains

Cited by in corpus (1)

• A New Artificial Neuron Proposal with Trainable Simultaneous Local and Global Activation Function