Understanding and Improving Graph Neural Networks with the Contextual Stochastic Block Model

Join a detailed webinar from the DEGAS series where Professor Ivan Dokmanić from the University of Basel explores the fundamentals and enhancements of graph neural networks through the lens of the contextual Stochastic Block Model (cSBM). Delve into advanced concepts and methodologies for optimizing neural network performance on graph-structured data, with practical insights on implementation and improvement strategies. Learn how the cSBM framework can be leveraged to better understand the behavior and limitations of graph neural networks while discovering innovative approaches to enhance their capabilities.