Search to aggregate neighborhood for graph neural network

TL;DR

This paper introduces SANE, a differentiable neural architecture search framework that automatically designs data-specific GNN architectures, improving efficiency and effectiveness over previous NAS methods for graph neural networks.

Contribution

It proposes a novel search space and a differentiable search algorithm tailored for GNN architecture search, addressing computational challenges.

Findings

SANE outperforms existing GNN models on multiple datasets.

SANE is more efficient than reinforcement learning-based NAS methods.

Experimental results demonstrate the effectiveness of the proposed approach.

Abstract

Recent years have witnessed the popularity and success of graph neural networks (GNN) in various scenarios. To obtain data-specific GNN architectures, researchers turn to neural architecture search (NAS), which has made impressive success in discovering effective architectures in convolutional neural networks. However, it is non-trivial to apply NAS approaches to GNN due to challenges in search space design and the expensive searching cost of existing NAS methods. In this work, to obtain the data-specific GNN architectures and address the computational challenges facing by NAS approaches, we propose a framework, which tries to Search to Aggregate NEighborhood (SANE), to automatically design data-specific GNN architectures. By designing a novel and expressive search space, we propose a differentiable search algorithm, which is more efficient than previous reinforcement learning based methods. Experimental results on four tasks and seven real-world datasets demonstrate the superiority of SANE compared to existing GNN models and NAS approaches in terms of effectiveness and efficiency. (Code is available at: https://github.com/AutoML-4Paradigm/SANE).