图神经网络及其变体研究综述

hrough a detailed reading and understanding of several papers, I have completed the writing of this review paper. I conducted in-depth analysis and comprehensive research on the fundamental concepts, core structures, and applications of Graph Neural Networks (GNN) and some of its variants, including Graph Convolutional Networks (GCN), GraphSAGE (Graph Sample and Aggregation), Graph Attention Networks (GAT), Graph Gated Neural Networks (GGNN), and Hierarchical Graph Neural Networks (HGNN) . The paper summarizes the research methods employed by the authors and provides insights into the functionalities and applications realized by the models they developed. Additionally, I present my understanding of the future development and research directions of GNN.