GraphPAS: Parallel Architecture Search for Graph Neural Networks

TL;DR

GraphPAS introduces a parallel search framework for GNN architectures that enhances efficiency and accuracy by sharing evolution learning and dynamic mutation strategies, enabling scalable exploration of large graph data.

Contribution

The paper presents a novel parallel architecture search method for GNNs that improves search efficiency and accuracy over existing approaches.

Findings

GraphPAS outperforms state-of-the-art models in accuracy.

The framework significantly reduces search time for large graphs.

Dynamic mutation based on entropy improves search effectiveness.

Abstract

Graph neural architecture search has received a lot of attention as Graph Neural Networks (GNNs) has been successfully applied on the non-Euclidean data recently. However, exploring all possible GNNs architectures in the huge search space is too time-consuming or impossible for big graph data. In this paper, we propose a parallel graph architecture search (GraphPAS) framework for graph neural networks. In GraphPAS, we explore the search space in parallel by designing a sharing-based evolution learning, which can improve the search efficiency without losing the accuracy. Additionally, architecture information entropy is adopted dynamically for mutation selection probability, which can reduce space exploration. The experimental result shows that GraphPAS outperforms state-of-art models with efficiency and accuracy simultaneously.