Just Jump: Dynamic Neighborhood Aggregation in Graph Neural Networks
This addresses the limitation of simple neighborhood aggregation in graph neural networks for researchers and practitioners in graph representation learning, though it appears incremental.
The paper tackles the problem of rigid neighborhood aggregation in graph neural networks by proposing a dynamic neighborhood aggregation procedure guided by attention, which allows selective and node-adaptive aggregation of neighboring embeddings. The approach demonstrates effectiveness in transductive node-classification experiments.
We propose a dynamic neighborhood aggregation (DNA) procedure guided by (multi-head) attention for representation learning on graphs. In contrast to current graph neural networks which follow a simple neighborhood aggregation scheme, our DNA procedure allows for a selective and node-adaptive aggregation of neighboring embeddings of potentially differing locality. In order to avoid overfitting, we propose to control the channel-wise connections between input and output by making use of grouped linear projections. In a number of transductive node-classification experiments, we demonstrate the effectiveness of our approach.