Do Not Train It: A Linear Neural Architecture Search of Graph Neural Networks
This addresses efficiency and accuracy issues in NAS for GNNs, offering a novel approach that is significantly faster and more accurate, though it is incremental in optimizing existing NAS methods.
The paper tackled the high computational cost and optimization difficulty in neural architecture search for graph neural networks (GNNs) by proposing a method that leverages GNNs' expressive power without training, resulting in state-of-the-art performance that is up to 200× faster and 18.8% more accurate than baselines.
Neural architecture search (NAS) for Graph neural networks (GNNs), called NAS-GNNs, has achieved significant performance over manually designed GNN architectures. However, these methods inherit issues from the conventional NAS methods, such as high computational cost and optimization difficulty. More importantly, previous NAS methods have ignored the uniqueness of GNNs, where GNNs possess expressive power without training. With the randomly-initialized weights, we can then seek the optimal architecture parameters via the sparse coding objective and derive a novel NAS-GNNs method, namely neural architecture coding (NAC). Consequently, our NAC holds a no-update scheme on GNNs and can efficiently compute in linear time. Empirical evaluations on multiple GNN benchmark datasets demonstrate that our approach leads to state-of-the-art performance, which is up to $200\times$ faster and $18.8\%$ more accurate than the strong baselines.