Deep Graph Mapper: Seeing Graphs through the Neural Lens (remote talk, with Cristian Bodnar)

Abstract

Recent advancements in graph representation learning have led to the emergence of condensed encodings that capture the main properties of a graph. However, even though these abstract representations are powerful for downstream tasks, they are not equally suitable for visualisation purposes. In this work, we merge Mapper, an algorithm from the field of Topological Data Analysis, with the expressive power of graph neural networks to produce hierarchical, topologically-grounded visualisations of graphs. These visualisations do not only help discern the structure of complex graphs, but also provide a means of understanding the models applied to them for solving various tasks. We further demonstrate the suitability of Mapper as a topological framework for graph pooling by showing an equivalence with minCUT and DiffPool. Building upon this framework, we introduce a novel pooling algorithm based on PageRank, which obtains competitive results with state-of-the-art methods on graph classification benchmarks. This talk has also been given on Tuesday, March 3rd 2020, at the AI Research Group Talks of the Department of Computer Science, University of Cambridge.

Dr Cătălina Cangea

Quantitative Researcher

Quantitative researcher with 9 years of ML experience, most recently co-lead of Generative Music at Google DeepMind, with a PhD from the University of Cambridge, and inhaler of music :) Motivated by contributing ML-based knowledge and improvements to real-world systems!