In the last few years, graph neural networks (GNNs) have emerged as a promising new supervised learning framework capable of bringing the power of deep representation learning to graph and relational data.
During The Web Conference in April, AWS deep learning scientists and engineers George Karypis, Zheng Zhang, Minjie Wang, Da Zheng, and Quan Gan presented a tutorial on GNNs.
The tutorial offers an overview of how learning GNNs can be used to solve problems such as detecting fraud and abuse (e.g., malicious accounts, fraudulent financial transactions, fake reviews), supporting customer recommendations (e.g., suggesting relevant products, jobs, articles, etc.), and delivering marketing campaigns (e.g., targeting who should get a discount, identifying influencers).
Watch the video presentation to learn more about putting GNNs to use in learning applications, and get an introduction and training on the AWS Deep Graph Library, a new software framework that simplifies the development of efficient GNN-based training and inference programs.
Tutorial sections:
- Overview of graph neural networks
- Overview of Deep Graph Library (DGL)
- GNN models for basic graph tasks
- GNN training on large graphs
- GNN models for real-world applications
