ELM-DeepONets: Backpropagation-Free Training of Deep Operator Networks via Extreme Learning Machines

TL;DR

ELM-DeepONets introduces a backpropagation-free training method for Deep Operator Networks using Extreme Learning Machines, significantly reducing computational costs while maintaining high accuracy in operator learning tasks.

Contribution

The paper presents a novel ELM-based training framework for DeepONets that eliminates backpropagation, enabling faster and more scalable operator learning.

Findings

Achieves high accuracy on nonlinear ODEs and PDEs

Reduces training time compared to traditional methods

Maintains performance without backpropagation

Abstract

Deep Operator Networks (DeepONets) are among the most prominent frameworks for operator learning, grounded in the universal approximation theorem for operators. However, training DeepONets typically requires significant computational resources. To address this limitation, we propose ELM-DeepONets, an Extreme Learning Machine (ELM) framework for DeepONets that leverages the backpropagation-free nature of ELM. By reformulating DeepONet training as a least-squares problem for newly introduced parameters, the ELM-DeepONet approach significantly reduces training complexity. Validation on benchmark problems, including nonlinear ODEs and PDEs, demonstrates that the proposed method not only achieves superior accuracy but also drastically reduces computational costs. This work offers a scalable and efficient alternative for operator learning in scientific computing.