PiNet: Attention Pooling for Graph Classification
This work addresses graph classification, a key problem in domains like chemo-informatics, but it appears incremental as it builds on existing graph neural network methods.
The authors tackled graph classification by proposing PiNet, an attention-based pooling mechanism, achieving superior performance in distinguishing isomorphic graph classes and competitive results on chemo-informatics datasets.
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.