Research map / Architectures and efficiency / Neural architecture search
Regularized Evolution for Image Classifier Architecture Search
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace worst-in-tournament removal with removing the oldest population member (ablation) | tournament-selection evolution · CIFAR-10 and other tested dataset/search-space contexts | search accuracy | Helped |
| remove age-based population replacement (ablation) | aging evolution · larger search spaces SP-II and SP-III | search accuracy | Hurt |
| remove age-based population replacement (ablation) | aging evolution · G-ImageNet/SP-I | search accuracy | No clear effect |
| use aging rather than non-aging evolution (ablation) | evolution on a noisy toy optimization task · higher-dimensional unit-cube search | simulated accuracy | Helped |
| use evolution instead of reinforcement learning | NASNet-space search · CIFAR-10, early search stages | search accuracy | Helped |
| use evolution instead of reinforcement learning | NASNet-space search · CIFAR-10, after 20k models | search accuracy | No clear effect |
| use evolution instead of random search | NASNet-space search · CIFAR-10 | augmented model test accuracy | Helped |
| increase AmoebaNet-A size from N=6, F=190 to N=6, F=448 | AmoebaNet-A · ImageNet | top-1 accuracy (%): 82.8 -> 83.9 | Helped |
| increase AmoebaNet-A size from N=6, F=190 to N=6, F=448 | AmoebaNet-A · ImageNet | top-5 accuracy (%): 96.1 -> 96.6 | Helped |
| increase AmoebaNet-A filters from F=32 to F=36 | AmoebaNet-A · CIFAR-10 test | test error (%): 3.40 ± 0.08 -> 3.34 ± 0.06 | Helped |
| change tournament sample size S across tested configurations (ablation) | aging evolution · CIFAR-10 architecture search | search performance | Mixed |
Closest papers by meaning
- Large-Scale Evolution of Image Classifiers (ICML 2017 · similarity 0.91)
- Hierarchical Representations for Efficient Architecture Search (ICLR 2018 · similarity 0.89)
- EG-NAS: Neural Architecture Search with Fast Evolutionary Exploration (AAAI 2024 · similarity 0.85)
- Runtime Analysis of Evolutionary NAS for Multiclass Classification (ICML 2025 · similarity 0.85)
- Neural Fine-Tuning Search for Few-Shot Learning (ICLR 2024 · similarity 0.84)
- Boosting Order-Preserving and Transferability for Neural Architecture Search: a Joint Architecture Refined Search and Fine-tuning Approach (CVPR 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.