Abstract:Machine learning has been successfully applied to various fields of scientific computing in recent years. In this work, we propose a sparse radial basis function neural network method to solve elliptic partial differential equations (PDEs) with multiscale coefficients. Inspired by the deep mixed residual method, we rewrite the second-order problem into a first-order system and employ multiple radial basis function neural networks (RBFNNs) to approximate unknown functions in the system. To aviod the overfitting due to the simplicity of RBFNN, an additional regularization is introduced in the loss function. Thus the loss function contains two parts: the $L_2$ loss for the residual of the first-order system and boundary conditions, and the $\ell_1$ regularization term for the weights of radial basis functions (RBFs). An algorithm for optimizing the specific loss function is introduced to accelerate the training process. The accuracy and effectiveness of the proposed method are demonstrated through a collection of multiscale problems with scale separation, discontinuity and multiple scales from one to three dimensions. Notably, the $\ell_1$ regularization can achieve the goal of representing the solution by fewer RBFs. As a consequence, the total number of RBFs scales like $\mathcal{O}(\varepsilon^{-n\tau})$, where $\varepsilon$ is the smallest scale, $n$ is the dimensionality, and $\tau$ is typically smaller than $1$. It is worth mentioning that the proposed method not only has the numerical convergence and thus provides a reliable numerical solution in three dimensions when a classical method is typically not affordable, but also outperforms most other available machine learning methods in terms of accuracy and robustness.

Functional Modelling of Large Scattered Data Sets Using Neural Networks

A Novel Nonlinear Dynamic System Modeling Approach Using Radial Basis Function Neural Networks

Parameterization of point-cloud freeform surfaces using adaptive sequential learning RBFnetworks

Modelling variability in dynamic functional brain networks using embeddings

Solving multiscale elliptic problems by sparse radial basis function neural networks

The Growing Radial Basis Function (RBF) Neural Network and Its Applications.

High-dimensional Functional Graphical Model Structure Learning via Neighborhood Selection Approach

Unravelled multilevel transformation networks for predicting sparsely observed spatio-temporal dynamics

Flexible Basis Representations for Modeling Large Non-Gaussian Spatial Data

Structured Radial Basis Function Network: Modelling Diversity for Multiple Hypotheses Prediction

Error-driven Active Learning in Growing Radial Basis Function Networks for Early Robot Learning

Adaptive point-cloud surface interpretation

Unraveled Multilevel Transformation Networks for Predicting Sparsely-Observed Spatiotemporal Dynamics

On-line Stable Nonlinear Modelling by Structurally Adaptive Neural Nets

Scalable Spatiotemporal Prediction with Bayesian Neural Fields

A Distributed Neural Hybrid System Learning Framework in Modeling Complex Dynamical Systems

A Functional Extension of Semi-Structured Networks

Recursive nearest neighbor co‐kriging models for big multi‐fidelity spatial data sets

A New Method for Decision on the Structure of RBF Neural Network

Robust Deep Neural Network Estimation for Multi-dimensional Functional Data

RBF Neural Networks Design with Graph Based Structural Information from Dominating Sets