Abstract:Graph structured data often possess dynamic characters in nature. Recent years have witnessed the increasing attentions paid to dynamic graph neural networks for modelling graph data. However, almost all existing approaches operate under the assumption that, upon the establishment of a new link, the embeddings of the neighboring nodes should undergo updates to learn temporal dynamics. Nevertheless, these approaches face the following limitation: If the node introduced by a new connection contains noisy information, propagating its knowledge to other nodes becomes unreliable and may even lead to the collapse of the model. In this paper, we propose Ada-DyGNN: a robust knowledge Adaptation framework via reinforcement learning for Dynamic Graph Neural Networks. In contrast to previous approaches, which update the embeddings of the neighbor nodes immediately after adding a new link, Ada-DyGNN adaptively determines which nodes should be updated. Considering that the decision to update the embedding of one neighbor node can significantly impact other neighbor nodes, we conceptualize the node update selection as a sequence decision problem and employ reinforcement learning to address it effectively. By this means, we can adaptively propagate knowledge to other nodes for learning robust node embedding representations. To the best of our knowledge, our approach constitutes the first attempt to explore robust knowledge adaptation via reinforcement learning specifically tailored for dynamic graph neural networks. Extensive experiments on three benchmark datasets demonstrate that Ada-DyGNN achieves the state-of-the-art performance. In addition, we conduct experiments by introducing different degrees of noise into the dataset, quantitatively and qualitatively illustrating the robustness of Ada-DyGNN.

Incorporating Self-Attentions into Robust Spatial-Temporal Graph Representation Learning Against Dynamic Graph Perturbations

Towards Robust Graph Convolution Network via Stochastic Activation

A Robust graph attention network with dynamic adjusted Graph

STGSN — A Spatial–Temporal Graph Neural Network framework for time-evolving social networks

Spatial-temporal Graph Convolutional Networks with Diversified Transformation for Dynamic Graph Representation Learning

Temporal Generalization Estimation in Evolving Graphs

Fast and Robust Contextual Node Representation Learning over Dynamic Graphs

A Spatial-Temporal Graph Mining Algorithm Based on Spatial-Temporal Sparse Attention

Improving Dynamic Graph Convolutional Network with Fine-Grained Attention Mechanism

STG-Mamba: Spatial-Temporal Graph Learning via Selective State Space Model

Spatial-Temporal Graph Learning with Adversarial Contrastive Adaptation

Sparse Graphs-based Dynamic Attention Networks

Long-term Spatio-temporal Forecasting Via Dynamic Multiple-Graph Attention

Neighborhood Extended Dynamic Graph Neural Network.

Spatio-Temporal Sparsification for General Robust Graph Convolution Networks

Dynamic Spatiotemporal Graph Neural Network with Tensor Network

Learning Robust Node Representations on Graphs.

Temporal Graph Transformer for Dynamic Network

Robust Knowledge Adaptation for Dynamic Graph Neural Networks

Dynamic Graph Representation Learning via Edge Temporal States Modeling and Structure-reinforced Transformer

A Recurrent Spatio-Temporal Graph Neural Network Based on Latent Time Graph for Multi-Channel Time Series Forecasting