Representation Learning with Graph Autoencoders and Applications to Music Recommendation

Explore graph autoencoders (GAE) and their applications in music recommendation in this comprehensive talk by Dr. Guillaume Salha-Galvan, research coordinator at Deezer. Delve into the speaker's Ph.D. research at École Polytechnique, focusing on overcoming critical limitations of GAE models for industrial adoption. Learn about advancements in scalability for large industrial graphs, solutions for directed and dynamic graphs, and real-world applications in graph-based music recommendation at Deezer. Gain insights into unsupervised node embedding methods and their potential for solving graph-based machine learning problems such as link prediction and community detection.

Representation Learning with Graph Autoencoders and Applications to Music Recommendation