PiNet: Attention Pooling for Graph Classification
arXiv:2008.04575
Abstract
We propose PiNet, a generalised differentiable attention-based pooling mechanism for utilising graph convolution operations for graph level classification. We demonstrate high sample efficiency and superior performance over other graph neural networks in distinguishing isomorphic graph classes, as well as competitive results with state of the art methods on standard chemo-informatics datasets.
4 pages, 3 figures 1 table
References in corpus (7)
- Simplifying Graph Convolutional Networks
- Learning Convolutional Neural Networks for Graphs
- Predicting multicellular function through multi-layer tissue networks
- MolGAN: An implicit generative model for small molecular graphs
- How Powerful are Graph Neural Networks?
- SoundNet: Learning Sound Representations from Unlabeled Video
- Janossy Pooling: Learning Deep Permutation-Invariant Functions for Variable-Size Inputs