Neural Architecture Search as Sparse Supernet

Authors

  • Yan Wu ETH Zurich
  • Aoming Liu ETH Zurich
  • Zhiwu Huang ETH Zurich
  • Siwei Zhang ETH Zurich
  • Luc Van Gool ETH Zurich VISICS, KU Leuven

Keywords:

Transfer/Adaptation/Multi-task/Meta/Automated Learning

Abstract

This paper aims at enlarging the problem of Neural Architecture Search (NAS) from Single-Path and Multi-Path Search to automated Mixed-Path Search. In particular, we model the NAS problem as a sparse supernet using a new continuous architecture representation with a mixture of sparsity constraints. The sparse supernet enables us to automatically achieve sparsely-mixed paths upon a compact set of nodes. To optimize the proposed sparse supernet, we exploit a hierarchical accelerated proximal gradient algorithm within a bi-level optimization framework. Extensive experiments on Convolutional Neural Network and Recurrent Neural Network search demonstrate that the proposed method is capable of searching for compact, general and powerful neural architectures.

Downloads

Published

2021-05-18

How to Cite

Wu, Y., Liu, A., Huang, Z., Zhang, S., & Van Gool, L. (2021). Neural Architecture Search as Sparse Supernet. Proceedings of the AAAI Conference on Artificial Intelligence, 35(12), 10379-10387. Retrieved from https://ojs.aaai.org/index.php/AAAI/article/view/17243

Issue

Section

AAAI Technical Track on Machine Learning V