Principal Investigator: Associate Professor He Bingsheng, SoC
Co-Principal Investigator: Associate Professor Wong Weng-Fai, SoC

Graph Neural Networks (GNNs), the neural network-based method on graph-structured data, has successfully generalized deep learning methods to model complex relationships and interdependencies on graphs and manifolds. It hybridizes the expressiveness of graphs and the learning process. In recent years，we have seen a significant growth in GNN research in academia and industry. They have achieved state-of-the-art performance in many tasks such as node classification, link prediction, and graph classification. A recent article offers a good summary of the emerging applications of GNN, including 1) Recommendation systems in Uber Eats, Alibaba, Amazon, Pinterest and etc. 2) Computer vision problems in Facebook, Magic Leap and etc. 3) Combinatorial optimization problems in many applications in finance, logistics, energy, life sciences, drug discovery and hardware design. For example, the Google Brain team used GNN to optimize the power, area, and performance of Google’s TPU.