Stochastic Graph Recurrent Neural Network
This work addresses the problem of predicting properties in dynamic networks for applications in fields like social networks or biology, representing an incremental improvement over static graph methods.
The paper tackles the challenge of modeling evolving graphs by proposing SGRNN, a neural architecture that uses stochastic latent variables to capture changes in node attributes and topology, demonstrating effectiveness through extensive experiments on real-world datasets.
Representation learning over graph structure data has been widely studied due to its wide application prospects. However, previous methods mainly focus on static graphs while many real-world graphs evolve over time. Modeling such evolution is important for predicting properties of unseen networks. To resolve this challenge, we propose SGRNN, a novel neural architecture that applies stochastic latent variables to simultaneously capture the evolution in node attributes and topology. Specifically, deterministic states are separated from stochastic states in the iterative process to suppress mutual interference. With semi-implicit variational inference integrated to SGRNN, a non-Gaussian variational distribution is proposed to help further improve the performance. In addition, to alleviate KL-vanishing problem in SGRNN, a simple and interpretable structure is proposed based on the lower bound of KL-divergence. Extensive experiments on real-world datasets demonstrate the effectiveness of the proposed model. Code is available at https://github.com/StochasticGRNN/SGRNN.