Neural Architecture Search

Table of contents

• 1. NAS introdction
• 2. NAS reviews
• 3. NAS Papers
  • Dataset
  • Search Space Definition
  • Evolutionary Algorithms
  • Reinforcement Learning
  • Others
• Application

Introduction

Neural Architecture Search can be seen as subfield of AutoML and has signicant overlap with hyper-parameter optimization and meta-learning. In NAS, We can divide it into EA(GA) + NAS, RL + NAS(and so on...I am going to summarize them).

Reviews

• 09.08, 2018 Neural Architecture Search: A Survey [paper]

Papers

This block is collated by Jingwen Pan(潘婧文). Thanks for her contributions! 🍰

Dataset

• 2019.02.25 NAS-Bench-101: Towards Reproducible Neural Architecture Search [paper] [code]

Search Space Definition

• 2018.12.27 ICLR 2019 Neural Architecture Search Over a Graph Search Space [paper]

Evolutionary Algorithms

• 2017.03.04 Evolving Deep Neural Networks [paper]

• 2017.03.04 ICCV'2017 Genetic CNN [paper] [code]

• 2017.06.11 ICML'2017 Large-Scale Evolution of Image Classifiers [paper]

• 2018.02.22 ICLR'2018 Hierarchical Representations for Efficient Architecture Search [paper]

• 2018.02.24 TEVC Evolving Unsupervised Deep Neural Networks for Learning Meaningful Representations [paper]

• 2018.06.04 ICLR'2018 PPP-Net: Platform-aware Progressive Search for Pareto-optimal Neural Architectures [paper]

• 2018.07.24 Efficient Multi-objective Neural Architecture Search via Lamarckian Evolution [paper]

• 2018.07.26 Progressive Neural Architecture Search [paper]

• 2018.10.08 GECCO'2019 NSGA-NET: A Multi-Objective Genetic Algorithm for Neural Architecture Search [paper]

• 2018.10.16 Evolutionary Stochastic Gradient Descent for Optimization of Deep Neural Networks [paper]

• 2018.10.26 Regularized Evolution for Image Classifier Architecture Search [paper]

Reinforcement Learning

• 2017.02.15 ICLR2017, Neural Architecture Search with Reinforcement Learning [paper]

• 2017.03.22 ICLR'2017 Designing Neural Network Architectures using Reinforcement Learning [paper]

• 2017.11.21 AAAI'2018 Efficient Architecture Search by Network Transformation [paper]

• 2018.02.12 Efficient Neural Architecture Search via Parameter Sharing [paper]

• 2018.04.11 Learning Transferable Architectures for Scalable Image Recognition [paper]

• 2018.05.14 CVPR'2018 Practical Block-wise Neural Network Architecture Generation [paper]

• 2018.06.27 JMLR MONAS: Multi-Objective Neural Architecture Search using Reinforcement Learning [paper]

Others

• 2017.04.28 ICCV'2017 DeepArchitect: Automatically Designing and Training Deep Architectures [paper] [code]

• 2018.06.10 Neural Architecture Search with Bayesian Optimisation and Optimal Transport [paper]

• 2018.08.30 ICCAD'18 Invited Paper Searching Toward Pareto-Optimal Device-Aware Neural Architectures [paper]

• 2018.09.05 NIPS'2018 Neural Architecture Optimization [paper]

• 2018.09.07 CVPR'2018 Reinforced Evolutionary Neural Architecture Search [paper]

• 2018.09.11 NIPS'2018 Searching for Efficient Multi-Scale Architectures for Dense Image Prediction [paper]

• 2018.10.10 Automatic Configuration of Deep Neural Networks with Parallel Efficient Global Optimization [paper]

• 2018.10.12 Graph Hypernetworks for Neural Architecture Search [paper]

• 2019.04.02 Exploring Randomly Wired Neural Networks for Image Recognition [paper] [code]

• 2019.5.23 Multinomial Distribution Learning for Effective Neural Architecture Search [paper] [code]

[MORE Paper]

Application

• 2019.01.24 Fast, Accurate and Lightweight Super-Resolution with Neural Architecture Search [paper]