Graphs are everywhere - but how can we reveal meaningful structure in graphs? Nowadays, in many real-world applications, the underlying data is modelled as graphs to represent complex objects and their relationships: cities in a road network, atoms in a molecule, friendships in social networks, connections in computer networks, and links among web pages. However, graphs in real-world applications are often highly complex, dynamic, and very large, which poses great challenges for analysing graphs. Traditional algorithms on graphs primarily aim to explore efficient algorithmic solutions to graph-related problems. In recent years, deep learning techniques such as graph neural networks have emerged to study graph-related problems from a machine learning perspective. In this talk, I will present some of our recent research work in these areas and discuss possible future directions to bridge traditional algorithms on graphs with graph learning.

Join Zoom Meeting