Abstract: Neural Operators extend traditional neural networks to learn mappings between infinite-dimensional function spaces. Graph Neural Operators (GNOs), which operate on geometric structures, have ...

Abstract: In this study, we proposed a new method combining graph sampling aggregation network (GraphSAGE), Contrastive Learning (CL) and Bayesian Optimization. First, GraphSAGE was used to model the ...