Large-Scale Graph Neural Network for Artificial Intelligence
- 1.2 Data science
- 1.3 Artificial intelligence fundamentals
- 2.4 High performance
- 3.4 Chemistry
- 3.5 Biology
- 3.8 Informatics
Toyotaro Suzumura
Graduate School of Information Science and Technology
Professor
A graph is a simple but fundamental data structure with which many real-world applications can be efficiently modelled, such as social network, purchase network, spatial-temporal network, biological network, material science, knowledge graph, and so forth. This project focuses on new learning methods for expressing graph structures using neural network called “Graph Neural Networks”.
Related links
Related publications
- EvolveGCN: Evolving Graph Convolutional Networks for Dynamic Graphs. AAAI 2020
- Efficient Scaling of Dynamic Graph Neural Networks, ACM/IEEE Supercomputing 2021
Contact
- Toyotaro Suzumura
- Email: suzumura[at]ds.itc.u-tokyo.ac.jp
※[at]=@
