Machine learningML & AI Methods

Generating synthetic data for neural operators

A novel method for creating synthetic functional training data for deep learning solutions to partial differential equations (PDEs) is proposed, eliminating the need for a numerical PDE solver and potentially broadening the scope for developing neural PDE solvers.