Research map / Label-efficient and robust learning / Noisy labels
Re-Active Learning: Active Learning with Relabeling
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 |
|---|---|---|---|
| use impactPLOPT(7) instead of uncertainty sampling without relabeling (USXU) | logistic regression, active learning · Synthetic Gaussian, 90 features, label accuracy 0.75 | accuracy: p < 0.001 | Helped |
| use impactPLOPT(7) instead of USXU | logistic regression · Internet Ads and Arrhythmia (synthetic labels) | accuracy: p < 0.001 | Helped |
| use impactPLOPT(7) instead of baselines | logistic regression · Galaxy Zoo (real labels) | F-score | No clear effect |
| use impactPLOPT(7) instead of baselines | logistic regression · Relation extraction (real labels) | F-score: p < 0.01 | Helped |
| add optimism (max impact) to impact sampling (ablation) | impactEXP(2) · Synthetic Gaussian, 90 features | accuracy | Helped |
| add pseudo-lookahead to optimistic impact sampling (ablation) | impactOPT(2) · Synthetic Gaussian, 90 features | accuracy | Helped |
| increase candidate pool from 2 to 7 points (ablation) | impactPLOPT · Synthetic Gaussian, 90 features | accuracy | Helped |
| increase alpha from 0.1 in USalpha_X (ablation) | uncertainty sampling with relabeling · Synthetic Gaussian, 90 features | accuracy | Hurt |
| apply optimism and pseudo-lookahead to EER (ablation) | EER · Synthetic Gaussian | accuracy | No clear effect |
| impact sampling instead of uncertainty sampling | active learning · Relation extraction | selection time: 7 hours vs 210 seconds | Hurt |
Closest papers by meaning
- STARS: Spatial-Temporal Active Re-sampling for Label-Efficient Learning from Noisy Annotations (AAAI 2023 · similarity 0.90)
- Instance-wise Supervision-level Optimization in Active Learning (CVPR 2025 · similarity 0.88)
- Towards Cost-Effective Learning: A Synergy of Semi-Supervised and Active Learning (CVPR 2025 · similarity 0.88)
- Active Learning From Positive and Unlabeled Examples (NeurIPS 2026 · similarity 0.87)
- Improved Algorithm for Deep Active Learning under Imbalance via Optimal Separation (ICML 2025 · similarity 0.86)
- A Cross-Domain Benchmark for Active Learning (NeurIPS 2024 · similarity 0.86)
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.