As the Complex Network Analysis (CNA) discussion group continues its activities, had the pleasure to host Jiong Zhu from the University of Michigan, to discuss about Limitations and effective designs for graph neural networks beyond homophily assumptions, online, on Wednesday 21 April 2021.
The CNA discussion group aims to bring together scientists from different groups and teams of the Institute that want to share research ideas, experimental results, expertise, and knowledge around the area of Complex Networks, probabilistic graphs, and related research topics. In this context, discussions and presentations take place twice a month.
You can also find this session video, but also previous presentations and videos, here.
Short bio: Jiong Zhu is a Ph.D. candidate in Computer Science and Engineering at the University of Michigan working with Danai Koutra. His research interest is on Graph Neural Network (GNN). Currently, he is working on understanding and improving GNNs on graphs with complex properties, such as graphs that go beyond the traditional homophily assumption (i.e., where the linked nodes are similar) by showing heterophily. He was a Master’s student in Electrical and Computer Engineering at UM. Before joining UM, he received his bachelor degree at Xi’an Jiaotong University, where he was a student of the Special Class for the Gifted Young.
If you wish to receive updates regarding CNA discussions, you can subscribe to firstname.lastname@example.org by sending an email to email@example.com with the subject: subscribe.