BETANAS: BalancEd TrAining and selective drop for Neural Architecture Search
This work addresses efficiency and fairness issues in neural architecture search for machine learning practitioners, representing an incremental improvement over existing methods.
The paper tackled the problem of conflicts and unfair comparisons in weight-sharing neural architecture search by proposing a balanced training strategy and selective drop mechanism, achieving a leading performance of 79.0% on ImageNet under mobile settings.
Automatic neural architecture search techniques are becoming increasingly important in machine learning area. Especially, weight sharing methods have shown remarkable potentials on searching good network architectures with few computational resources. However, existing weight sharing methods mainly suffer limitations on searching strategies: these methods either uniformly train all network paths to convergence which introduces conflicts between branches and wastes a large amount of computation on unpromising candidates, or selectively train branches with different frequency which leads to unfair evaluation and comparison among paths. To address these issues, we propose a novel neural architecture search method with balanced training strategy to ensure fair comparisons and a selective drop mechanism to reduce conflicts among candidate paths. The experimental results show that our proposed method can achieve a leading performance of 79.0% on ImageNet under mobile settings, which outperforms other state-of-the-art methods in both accuracy and efficiency.