Skip to main content

Trustworthy Graph Neural Networks (GNNs)

Primary supervisor

Shirui Pan

Graph machine learning, graph neural networks, in particular, is the frontier of deep learning. There has been an exponential growth of research on graph neural networks (GNNs) in the last few years, mainly focusing on how to develop accurate GNN models. The trustworthiness of GNNs is less considered. In this project, we will explore how to develop trustworthy GNN models. The following key aspects will be taken into consideration when developing GNN models.

  • Robustness of GNN models.
  • Explainability of GNN models.
  • Fairness of GNN models.
  • Privacy of GNN models.


You will be working with world-leading researchers in the area of GNNs. Potential outcomes of this project include 

  • Publications in KDD, ICML, NeurIPS, ICLR, AAAI.
  • Systems that can be deployed in real-life applications.


Our comprehensive survey on trustworthy GNNs can be found at


We will provide you with personalised training, job-ready skills, and opportunities to collaborate with world-leading researchers in the world and industry partners.



Required knowledge


Deep learning


Project funding

Project based scholarship

Learn more about minimum entry requirements.