Improved generalization with deep neural operators for engineering systems: Path towards digital twin

arXiv:2301.06701v3 Announce Type: replace
Abstract: Neural Operator Networks (ONets) represent a novel advancement in machine learning algorithms, offering a robust and generalizable alternative for approximating partial differential equations (PDEs) solutions. Unlike traditional Neural Networks (NN), which directly approximate functions, ONets specialize in approximating mathematical operators, enhancing their efficacy in addressing complex PDEs. In this work, we evaluate the capabilities of Deep Operator Networks (DeepONets), an ONets implementation using a branch/trunk architecture. Three test cases are studied: a system of ODEs, …

abstract advancement algorithms alternative arxiv cs.lg differential digital digital twin engineering functions machine machine learning machine learning algorithms networks neural networks novel operators path robust solutions stat.ap stat.co stat.ml systems twin type