Simplifying Architecture Search for Graph Neural Network
This work addresses the challenge of automating architecture design for GNNs, which is incremental as it builds on prior NAS methods to enhance efficiency and capability in a domain-specific context.
The paper tackles the problem of inefficient and limited expressive capability in neural architecture search (NAS) for Graph Neural Networks (GNNs) by proposing the SNAG framework, which includes a novel search space and a reinforcement learning-based search algorithm, achieving improved performance over existing methods like GraphNAS and Auto-GNN on real-world datasets.
Recent years have witnessed the popularity of Graph Neural Networks (GNN) in various scenarios. To obtain optimal data-specific GNN architectures, researchers turn to neural architecture search (NAS) methods, which have made impressive progress in discovering effective architectures in convolutional neural networks. Two preliminary works, GraphNAS and Auto-GNN, have made first attempt to apply NAS methods to GNN. Despite the promising results, there are several drawbacks in expressive capability and search efficiency of GraphNAS and Auto-GNN due to the designed search space. To overcome these drawbacks, we propose the SNAG framework (Simplified Neural Architecture search for Graph neural networks), consisting of a novel search space and a reinforcement learning based search algorithm. Extensive experiments on real-world datasets demonstrate the effectiveness of the SNAG framework compared to human-designed GNNs and NAS methods, including GraphNAS and Auto-GNN.