Professor Jia Cai was invited to give a lecture

发布时间: 2023-03-20 00:47:56

Topic：Graph Neural Networks: Algorithms and Theory

Speaker：Jia Cai  Guangdong University of Finance & Economics

Time：2023-03-13 15:30-16:30

Location：Academic hall 111

Introduction：

In real life, there are many scenarios where data is defined as graph data with complex topological structures. Graph data naturally represents the data structure in real life. However, the local structure of each node in the graph data varies. The diversity of these graphs brings new challenges to the design of graph neural network models. At the same time, in the era of big data, many graphs in practical applications contain millions or even tens of millions of nodes. This report will provide relevant algorithms and theoretical analysis for several core issues related to graph convolution neural networks: the theory of shallow graph neural networks in semi supervised multi classification situations, and multiscale graph convolution neural networks.