Relational Pooling for Graph Representations
Ryan Murphy · Balasubramaniam Srinivasan · Vinayak A Rao · Bruno Ribeiro

Wed Jun 12th 04:30 -- 04:35 PM @ Room 102

This work generalizes graph neural networks (GNNs) beyond those based on the Weisfeiler-Lehman (WL) algorithm, graph Laplacians, and graph diffusion kernels. Our approach, denoted Relational Pooling (RP), draws from the theory of finite partial exchangeability to provide a framework with maximal representation power for graphs. RP can work with existing graph representation models, and somewhat counterintuitively, can make them even more powerful than the original WL isomorphism test. Additionally, RP is the first theoretically sound framework to use architectures like Recurrent Neural Networks and Convolutional Neural Networks for graph classification. RP also has graph kernels as a special case. We demonstrate improved performance of novel RP-based graph representations over current state-of-the-art methods on a number of tasks.

Author Information

Ryan Murphy (Purdue University)

I am a Ph.D. candidate at Purdue University working with Professors Bruno Ribeiro (Computer Science) and Vinayak Rao (Statistics and Computer Science) on machine learning methods for graphs (networks) and unordered sequences. Applications for this work include predicting molecular properties, diagnosing neurodegenerative disease, and anomaly detection.

Balasubramaniam Srinivasan (Purdue University)
Vinayak A Rao (Purdue University)
Bruno Ribeiro (Purdue University)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors