Graph Neural Networks with Parallel Neighborhood Aggregations for Graph Classification

TL;DR

This paper introduces a parallel neighborhood aggregation GNN model called PA-GNN, which precomputes node features for faster graph classification, and proves its theoretical power comparable to the Weisfeiler-Lehman test, with state-of-the-art results.

Contribution

It develops a new PA-GNN model with theoretical guarantees and demonstrates its effectiveness and efficiency in graph classification tasks.

Findings

PA-GNN achieves state-of-the-art performance on multiple datasets.

Theoretical conditions ensure PA-GNN's discriminative power matches WL test.

Precomputing features reduces training and inference time.

Abstract

We focus on graph classification using a graph neural network (GNN) model that precomputes the node features using a bank of neighborhood aggregation graph operators arranged in parallel. These GNN models have a natural advantage of reduced training and inference time due to the precomputations but are also fundamentally different from popular GNN variants that update node features through a sequential neighborhood aggregation procedure during training. We provide theoretical conditions under which a generic GNN model with parallel neighborhood aggregations (PA-GNNs, in short) are provably as powerful as the well-known Weisfeiler-Lehman (WL) graph isomorphism test in discriminating non-isomorphic graphs. Although PA-GNN models do not have an apparent relationship with the WL test, we show that the graph embeddings obtained from these two methods are injectively related. We then propose a specialized PA-GNN model, called SPIN, which obeys the developed conditions. We demonstrate via numerical experiments that the developed model achieves state-of-the-art performance on many diverse real-world datasets while maintaining the discriminative power of the WL test and the computational advantage of preprocessing graphs before the training process.