Graph Attention Network
Graph Attention Networks (GATs) are a type of graph neural network designed to learn relationships within graph-structured data by assigning attention weights to different nodes based on their relevance to a given task. Current research focuses on improving GAT performance in various applications, including developing more sophisticated attention mechanisms, addressing limitations in handling heterophilic graphs and long-range dependencies, and integrating GATs with other techniques like transformers and multi-modal data fusion. This work has significant implications across diverse fields, enabling advancements in areas such as drug discovery, financial modeling, image analysis, and social network analysis through improved data representation and predictive modeling.
Papers
Heterogeneous Graph Reinforcement Learning for Dependency-aware Multi-task Allocation in Spatial Crowdsourcing
Yong Zhao, Zhengqiu Zhu, Chen Gao, En Wang, Jincai Huang, Fei-Yue Wang
A Heterogeneous Network-based Contrastive Learning Approach for Predicting Drug-Target Interaction
Junwei Hu, Michael Bewong, Selasi Kwashie, Wen Zhang, Vincent M. Nofong, Guangsheng Wu, Zaiwen Feng
GATher: Graph Attention Based Predictions of Gene-Disease Links
David Narganes-Carlon, Anniek Myatt, Mani Mudaliar, Daniel J. Crowther
Dual Stream Graph Transformer Fusion Networks for Enhanced Brain Decoding
Lucas Goene, Siamak Mehrkanoon
Graph Network Models To Detect Illicit Transactions In Block Chain
Hrushyang Adloori, Vaishnavi Dasanapu, Abhijith Chandra Mergu