Fast and Deep Graph Neural Networks
This addresses efficiency issues for researchers and practitioners in graph machine learning, though it is incremental as it builds on existing GNN and recurrent network ideas.
The paper tackles the efficiency problem in constructing deep graph neural networks (GNNs) by representing graphs as fixed points of a dynamical system using recurrent neural networks with untrained weights under a stability condition, achieving or improving state-of-the-art performance on graph classification tasks.
We address the efficiency issue for the construction of a deep graph neural network (GNN). The approach exploits the idea of representing each input graph as a fixed point of a dynamical system (implemented through a recurrent neural network), and leverages a deep architectural organization of the recurrent units. Efficiency is gained by many aspects, including the use of small and very sparse networks, where the weights of the recurrent units are left untrained under the stability condition introduced in this work. This can be viewed as a way to study the intrinsic power of the architecture of a deep GNN, and also to provide insights for the set-up of more complex fully-trained models. Through experimental results, we show that even without training of the recurrent connections, the architecture of small deep GNN is surprisingly able to achieve or improve the state-of-the-art performance on a significant set of tasks in the field of graphs classification.