Neural Architecture Search
Neural architecture search (NAS) is a technique for automating the design of artificial neural networks (ANN), a widely used model in the field of machine learning. NAS essentially takes the process of a human manually tweaking a neural network and learning what works well, and automates this task to discover more complex architectures.
Image Credit : NAS with Reinforcement Learning
Papers
Showing 1–10 of 1915 papers
All datasetsImageNetNAS-Bench-201, ImageNet-16-120CIFAR-10NAS-Bench-201, CIFAR-100NAS-Bench-201, CIFAR-10CIFAR-10 Image ClassificationCIFAR-100NATS-Bench Topology, CIFAR-10NATS-Bench Topology, CIFAR-100NATS-Bench Topology, ImageNet16-120Food-101NAS-Bench-101
Benchmark Results
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | Balanced Mixture | Accuracy (% ) | 91.55 | — | Unverified |
| 2 | GDAS | Top-1 Error Rate | 3.4 | — | Unverified |
| 3 | Bonsai-Net | Top-1 Error Rate | 3.35 | — | Unverified |
| 4 | Net2 (2) | Top-1 Error Rate | 3.3 | — | Unverified |
| 5 | μDARTS | Top-1 Error Rate | 3.28 | — | Unverified |
| 6 | NN-MASS- CIFAR-C | Top-1 Error Rate | 3.18 | — | Unverified |
| 7 | NN-MASS- CIFAR-A | Top-1 Error Rate | 3 | — | Unverified |
| 8 | DARTS (first order) | Top-1 Error Rate | 3 | — | Unverified |
| 9 | NASGEP | Top-1 Error Rate | 2.82 | — | Unverified |
| 10 | AlphaX-1 (cutout NASNet) | Top-1 Error Rate | 2.82 | — | Unverified |