论文标题
pinet：图形分类的注意集合
PiNet: Attention Pooling for Graph Classification
论文作者
论文摘要
我们提出了Pinet，这是一种基于广义的基于注意的集合机制，用于利用图形卷积操作进行图形分类。在区分同构图类别方面，我们证明了高样本效率和优于其他图神经网络的性能，以及具有标准化学信息形式数据集的最先进方法的竞争结果。
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.
