Overview
Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Explore the latest advancements in Graph Machine Learning with this comprehensive talk on Deep Graph Library (DGL) 1.0. Delve into the key features of this milestone release, including DGL Sparse, a new sparse matrix programming abstraction designed for cutting-edge Graph Neural Network (GNN) models. Learn about the advantages of DGL, its release history, and architecture overview. Gain insights into GNN research at ASAIL, focusing on scalability and improving GNNs through an energy unfolding perspective. Discover the future directions of graph machine learning and participate in an informative Q&A session. This 1 hour 15 minute presentation by Minjie Wang from Valence Labs offers valuable knowledge for both academic researchers and industry professionals working with large-scale, real-world graphs.
Syllabus
Intro & Outline
- What is DGL
- Advantages of DGL
- DGL Release History and Architecture Overview
- GNN Research at ASAIL - On Scalability
- Improving GNN with Energy Unfolding Perspective
- A Series of Works in This Direction
- Looking Forward
- Q+A
Taught by
Valence Labs