Sample-Efficient Neural Architecture Search by Learning Actions for Monte Carlo Tree Search
- 7 April 2021
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. PP (01628828) , 1
Abstract
Neural Architecture Search (NAS) has emerged as a promising technique for automatic neural network design. However, existing MCTS based NAS approaches often utilize manually designed action space, which is not directly related to the performance metric to be optimized (e.g., accuracy), leading to sample-inefficient explorations of architectures. To improve the sample efficiency, this paper proposes Latent Action Neural Architecture Search (LaNAS), which learns actions to recursively partition the search space into good or bad regions that contain networks with similar performance metrics. During the search phase, as different action sequences lead to regions with different performance, the search efficiency can be significantly improved by biasing towards the good regions. On three NAS tasks, empirical results demonstrate that LaNAS is at least an order more sample efficient than baseline methods including evolutionary algorithms, Bayesian optimizations, and random search. When applied in practice, both one-shot and regular LaNAS consistently outperform existing results. Particularly, LaNAS achieves 99.0% accuracy on CIFAR-10 and 80.8% top1 accuracy at 600 MFLOPS on ImageNet in only 800 samples, significantly outperforming AmoebaNet with 33x fewer samples. Our code is publicly available at https://github.com/facebookresearch/LaMCTS.Keywords
This publication has 27 references indexed in Scilit:
- Mastering the game of Go with deep neural networks and tree searchNature, 2016
- From Bandits to Monte-Carlo Tree Search: The Optimistic Principle Applied to Optimization and PlanningFoundations and Trends in Machine Learning, 2014
- An evaluation of sequential model-based optimization for expensive blackbox functions2013
- Optimistic planning for continuous-action deterministic systems2013
- Sequential Model-Based Optimization for General Algorithm Configuration2011
- Concave-Convex Adaptive Rejection SamplingJournal of Computational and Graphical Statistics, 2011
- An informational approach to the global optimization of expensive-to-evaluate functionsJournal of Global Optimization, 2008
- A rejection technique for sampling from T -concave distributionsACM Transactions on Mathematical Software, 1995
- Adaptive Rejection Metropolis Sampling within Gibbs SamplingJournal of the Royal Statistical Society Series C: Applied Statistics, 1995
- Letters to the EditorThe American Statistician, 1991