Temporal Graph
Temporal graphs represent dynamic relationships between entities evolving over time, aiming to model and predict these changes. Current research focuses on developing sophisticated neural network architectures, such as graph convolutional networks (GCNs) and transformers, often incorporating mechanisms like attention and recurrent layers to capture complex spatio-temporal dependencies. These advancements are improving the accuracy of predictions in diverse applications, including traffic forecasting, disease outbreak prediction, and even medical diagnosis using electronic health records. The field is also actively exploring efficient training methods and developing standardized benchmarks for evaluating model performance across various datasets and tasks.
Papers
diffIRM: A Diffusion-Augmented Invariant Risk Minimization Framework for Spatiotemporal Prediction over Graphs
Zhaobin Mo, Haotian Xiang, Xuan Di
SepsisCalc: Integrating Clinical Calculators into Early Sepsis Prediction via Dynamic Temporal Graph Construction
Changchang Yin, Shihan Fu, Bingsheng Yao, Thai-Hoang Pham, Weidan Cao, Dakuo Wang, Jeffrey Caterino, Ping Zhang