Timezone: »
We introduce Automorphism-based graph neural networks (Autobahn), a new family of graph neural networks. In an Autobahn, we decompose the graph into a collection of subgraphs and apply local convolutions that are equivariant to each subgraph's automorphism group. Specific choices of local neighborhoods and subgraphs recover existing architectures such as message passing neural networks. Our formalism also encompasses novel architectures: as an example, we introduce a graph neural network that decomposes the graph into paths and cycles. The resulting convolutions reflect the natural way that parts of the graph can transform, preserving the intuitive meaning of convolution without sacrificing global permutation equivariance. We validate our approach by applying Autobahn to molecular graphs, where it achieves results competitive with state-of-the-art message passing algorithms.
Author Information
Erik Thiede (Flatiron Institute)
Wenda Zhou (Columbia University)
Risi Kondor (Flatiron Institute)
Risi Kondor joined the Flatiron Institute in 2019 as a Senior Research Scientist with the Center for Computational Mathematics. Previously, Kondor was an Associate Professor in the Department of Computer Science, Statistics, and the Computational and Applied Mathematics Initiative at the University of Chicago. His research interests include computational harmonic analysis and machine learning. Kondor holds a Ph.D. in Computer Science from Columbia University, an MS in Knowledge Discovery and Data Mining from Carnegie Mellon University, and a BA in Mathematics from the University of Cambridge. He also holds a diploma in Computational Fluid Dynamics from the Von Karman Institute for Fluid Dynamics and a diploma in Physics from Eötvös Loránd University in Budapest.
More from the Same Authors
-
2021 : ATOM3D: Tasks on Molecules in Three Dimensions »
Raphael Townshend · Martin Vögele · Patricia Suriana · Alex Derry · Alexander Powers · Yianni Laloudakis · Sidhika Balachandar · Bowen Jing · Brandon Anderson · Stephan Eismann · Risi Kondor · Russ Altman · Ron Dror -
2022 : Multiresolution Mesh Networks For Learning Dynamical Fluid Simulations »
Bach Nguyen · Truong Son Hy · Long Tran-Thanh · Risi Kondor -
2022 : Predicting Drug-Drug Interactions using Deep Generative Models on Graphs »
Khang Ngo · Truong Son Hy · Risi Kondor -
2022 : A code superoptimizer through neural Monte-Carlo tree search »
Wenda Zhou · Olga Solodova · Ryan Adams -
2022 : A code superoptimizer through neural Monte-Carlo tree search »
Wenda Zhou · Olga Solodova · Ryan Adams -
2021 : ATOM3D: Tasks on Molecules in Three Dimensions »
Raphael Townshend · Martin Vögele · Patricia Suriana · Alex Derry · Alexander Powers · Yianni Laloudakis · Sidhika Balachandar · Bowen Jing · Brandon Anderson · Stephan Eismann · Risi Kondor · Russ Altman · Ron Dror -
2020 Tutorial: (Track2) Equivariant Networks Q&A »
Risi Kondor · Taco Cohen -
2020 Tutorial: (Track2) Equivariant Networks »
Risi Kondor · Taco Cohen -
2019 : Solutions »
Fitzroy Christian · Lily Hu · Risi Kondor · Brandeis Marshall · Fabian Rogers · Friederike Schuur · Emanuel Moss -
2019 Workshop: Minding the Gap: Between Fairness and Ethics »
Igor Rubinov · Risi Kondor · Jack Poulson · Manfred K. Warmuth · Emanuel Moss · Alexa Hagerty -
2019 Poster: Cormorant: Covariant Molecular Neural Networks »
Brandon Anderson · Truong Son Hy · Risi Kondor -
2019 Spotlight: Cormorant: Covariant Molecular Neural Networks »
Brandon Anderson · Truong Son Hy · Risi Kondor -
2019 Poster: Discrete Object Generation with Reversible Inductive Construction »
Ari Seff · Wenda Zhou · Farhan Damani · Abigail Doyle · Ryan Adams -
2018 Poster: Clebsch–Gordan Nets: a Fully Fourier Space Spherical Convolutional Neural Network »
Risi Kondor · Zhen Lin · Shubhendu Trivedi