How Powerful Are Graph Neural Networks?

Machine learning on graphs is an important and ubiquitous task with applications ranging from drug design to friendship recommendation in social networks. The primary challenge in this domain is finding a way to represent, or encode, graph structure so that it can be easily exploited by machine learning models. In this talk he will discuss methods that automatically learn to encode graph structure into lowdimensional embeddings, using techniques based on deep learning and nonlinear dimensionality reduction. He will provide a conceptual review of key advancements in this area of representation learning on graphs, including graph convolutional networks and their representational power.