NeurIPS 2018 accepted papers by research area
1,008 papers were accepted at NeurIPS 2018 (+320 vs 2017). Here is how they split across 20 research areas, and which topics gained the most ground.
See NeurIPS 2018 in the interactive map
Submitting to NeurIPS? Checklist →
Research areas
| Area | Papers | Share | Change in share vs 2017 |
|---|---|---|---|
| Learning theory and algorithms | 206 | 20.4% | -6.9 pts |
| Probabilistic and scientific modeling | 164 | 16.3% | -2.2 pts |
| Optimization | 113 | 11.2% | +0.9 pts |
| Architectures and efficiency | 96 | 9.5% | +1.2 pts |
| Label-efficient and robust learning | 73 | 7.2% | +0.6 pts |
| Reinforcement learning | 70 | 6.9% | +1.7 pts |
| Bandits and online learning | 43 | 4.3% | -1.8 pts |
| Image generation | 37 | 3.7% | -0.5 pts |
| Adversarial robustness and security | 33 | 3.3% | +2.5 pts |
| Graph learning | 32 | 3.2% | +0.4 pts |
| Detection and segmentation | 27 | 2.7% | +1.5 pts |
| Federated learning and privacy | 18 | 1.8% | +0.8 pts |
| Vision-language and multimodal | 16 | 1.6% | steady |
| Multi-agent systems and games | 15 | 1.5% | -0.4 pts |
| 3D vision | 15 | 1.5% | +0.8 pts |
| Video | 15 | 1.5% | +0.3 pts |
| Time series | 15 | 1.5% | +0.3 pts |
| Large language models | 9 | 0.9% | +0.6 pts |
| Continual learning | 6 | 0.6% | steady |
| Point clouds | 5 | 0.5% | +0.2 pts |
Fastest-growing topics at NeurIPS 2018
Change in share of NeurIPS papers from 2017 to 2018, topics with at least 10 papers.
- Certified and empirical robustness (Adversarial robustness and security): 4 → 24 papers
- RNNs and normalization (Architectures and efficiency): 31 → 55 papers
- Implicit bias and generalization (Optimization): 8 → 21 papers
- Open-set and few-shot detection (Detection and segmentation): 2 → 11 papers
- Distributed and compressed (Optimization): 7 → 18 papers
- Differential privacy (Federated learning and privacy): 5 → 15 papers
- Domain adaptation and generalization (Label-efficient and robust learning): 4 → 13 papers
- Causal discovery (Probabilistic and scientific modeling): 5 → 14 papers
- Policy gradient and actor-critic (Reinforcement learning): 16 → 30 papers
- CNN features (Detection and segmentation): 4 → 12 papers
More venues
Same year: ICML 2018 · ICLR 2018 · CVPR 2018 · AAAI 2018 · ECCV 2018 · TPAMI 2018 · IJCV 2018 · JMLR 2018
NeurIPS in other years: 2016 · 2017 · 2019 · 2020 · 2021 · 2022 · 2023 · 2024 · 2025 · 2026
