Research map / Learning theory and algorithms / Combinatorial and search optimization
Low-Budget Active Learning via Wasserstein Distance: An Integer Programming Approach
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 |
|---|---|---|---|
| Minimize Wasserstein distance with GBD and enhanced optimality cuts | active learning with SimCLR features · STL-10, B=40 | accuracy: 56.5 ± 1.2 -> 60.4 ± 1.5 | Helped |
| Minimize Wasserstein distance with GBD and enhanced optimality cuts | active learning with SimCLR features · CIFAR-10, B=10 | accuracy: 45.6 ± 5.9 -> 50.7 ± 3.0 | Helped |
| Minimize Wasserstein distance with GBD and enhanced optimality cuts | active learning with SimCLR features · CIFAR-10, B=40 | accuracy: 74.7 ± 4.3 -> 73.5 ± 1.6 | Hurt |
| Minimize Wasserstein distance with GBD and enhanced optimality cuts | active learning with SimCLR features · SVHN, B=40 | accuracy: 58.6 ± 6.3 -> 63.8 ± 6.1 | Helped |
| Add pruning constraints to Wass. + EOC (ablation) | active learning with SimCLR features · STL-10, B=10 | accuracy: 43.1 ± 0.7 -> 41.4 ± 3.2 | Hurt |
| Add pruning constraints to Wass. + EOC (ablation) | active learning with SimCLR features · CIFAR-10, B=10 | accuracy: 50.7 ± 3.0 -> 46.8 ± 1.6 | Hurt |
| Add pruning constraints to Wass. + EOC (ablation) | active learning with SimCLR features · SVHN, B=40 | accuracy: 63.2 ± 6.5 -> 63.8 ± 6.1 | Helped |
| Use Wasserstein optimization instead of k-medoids | core-set selection · SVHN, B=20 | Wasserstein objective: 0.172 -> 0.093 | Helped |
Closest papers by meaning
- Active Learning with Neural Networks: Insights from Nonparametric Statistics (NeurIPS 2022 · similarity 0.86)
- Active Learning of Deep Neural Networks via Gradient-Free Cutting Planes (ICML 2025 · similarity 0.86)
- Active Learning for Convolutional Neural Networks: A Core-Set Approach (ICLR 2018 · similarity 0.86)
- Active Learning Through a Covering Lens (NeurIPS 2022 · similarity 0.86)
- One-shot Active Learning Based on Lewis Weight Sampling for Multiple Deep Models (ICLR 2024 · similarity 0.85)
- Neural Active Learning Beyond Bandits (ICLR 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.