Artificial Intelligence
神经算子
Neural OperatorsA class of deep learning architectures designed to learn mappings between function spaces, enabling the solution of partial differential equations (PDEs) and the modeling of physical systems without discretizing the underlying domain. / 一类深度学习架构，旨在学习函数空间之间的映射，无需对底层域进行离散化即可求解偏微分方程（PDE）和建模物理系统。
- Knowledge type
- 数学理论
- First proposed or discovered
- Proposer
- Turkey Farm member
- Creator
- 农场主
- Last updated
- 2026-10-28 19:02:26
Sub-knowledge points
Physics-Informed Neural Networks (PINNs) / 数学理论Operator LearningDeepONetFourier Neural Operator (FNO)
Higher-order knowledge points
Main parameters and performance indicators
Physics-informed: Incorporates physical laws into the loss functionFourier Neural Operator (FNO): A specific architecture using Fourier transformsDeepONet: Branch and trunk network architecture
Open problems
Generalization to out-of-distribution dataComputational efficiency for real-time applicationsInterpretability of learned operators
Related literature
ranked by community starsNeural operators for power systems: A physics-informed framework for modeling power system componentsOpen source
Related products and links
No item recorded yet
Creation and modification trail
1 recordsRecord title: Neural Operators.
Node update feedback
review required