GRASP-GCN: Graph-Shape Prioritization for Neural Architecture Search under Distribution Shifts

TL;DR

This paper introduces GRASP-GCN, a graph convolutional network that enhances neural architecture prediction performance across different datasets and distribution shifts, reducing computational costs and improving generalization.

Contribution

The paper proposes GRASP-GCN, a novel ranking GCN that incorporates layer shape information to improve NAS prediction accuracy under distribution shifts.

Findings

Improves prediction accuracy by 3.3% on CIFAR-10.

Enhances generalization under data distribution shifts.

Creates a new NAS benchmark across four datasets.

Abstract

Neural Architecture Search (NAS) methods have shown to output networks that largely outperform human-designed networks. However, conventional NAS methods have mostly tackled the single dataset scenario, incuring in a large computational cost as the procedure has to be run from scratch for every new dataset. In this work, we focus on predictor-based algorithms and propose a simple and efficient way of improving their prediction performance when dealing with data distribution shifts. We exploit the Kronecker-product on the randomly wired search-space and create a small NAS benchmark composed of networks trained over four different datasets. To improve the generalization abilities, we propose GRASP-GCN, a ranking Graph Convolutional Network that takes as additional input the shape of the layers of the neural networks. GRASP-GCN is trained with the not-at-convergence accuracies, and improves the state-of-the-art of 3.3 % for Cifar-10 and increasing moreover the generalization abilities under data distribution shift.