Title:  Revisit Framelet-based Graph Neural Networks

Abstract:

In this talk, I will give an overview on our recent works in framelets applications in graph neural networks (GNN).  We decompose an input graphs into low-pass and high-pass frequencies components for network training, which then defines a framelet-based graph convolution.  The graph neural networks with the proposed framelet convolution and pooling achieve state-of-the-art performance in many node and graph prediction tasks. I will also explore how more general framelet-based GNN can be constructed and conditions under which the framelet-based GCN can avoid the oversmoothing issues with the conventional GNNs. Further I will also touch how the idea of framelet-based GNN can be extended for directed graphs.

Short Bio:

Junbin Gao is Professor of Big Data Analytics in the University of Sydney Business School at the University of Sydney and was a Professor in Computer Science in the School of Computing and Mathematics at Charles Sturt University, Australia. Professor Junbin Gao obtained his PhD from Dalian University of Technology, China in 1991.  Professor Junbin Gao is an International Standard expert in Machine Learning (a top 2% researcher in AI according to the 2022 Mendeley list from SRI). His current research interests cover Statistical Machine Learning, Bayesian Inference, Computer Vision and Image Analysis, Data Mining and Big Data including tensorial data, Numerical Optimization, and Visualization.