Research map / Label-efficient and robust learning / Recognition and long-tail
Self-Paced Active Learning: Query the Right Thing at the Right Time
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 |
|---|---|---|---|
| add easiness weighting (self-paced) on top of informative+representative selection: SPAL vs BMDR | SPAL · krvskp, 5% initial labels | accuracy: 0.937 -> 0.964 | Helped |
| SPAL vs BMDR | SPAL · thyroid, 5% initial labels | accuracy: 0.889 -> 0.929 | Helped |
| SPAL vs random selection | SPAL · tictactoe, 5% initial labels | accuracy: 0.748 -> 0.786 | Helped |
| SPAL vs ASPL (pseudo-labelling) | SPAL · gisette, 5% initial labels | accuracy: 0.933 -> 0.947 | Helped |
| SPAL vs ASPL | SPAL · phishing, 5% initial labels | accuracy: 0.935 -> 0.937 | No clear effect |
| SPAL vs ASPL, overall paired t-tests | SPAL · 9 datasets, win/tie/loss | accuracy: 18/27/0 win/tie/loss | Mixed |
| SPAL vs Random, overall paired t-tests | SPAL · 9 datasets, win/tie/loss | accuracy: 40/5/0 win/tie/loss | Helped |
| SPAL vs BMDR, overall paired t-tests | SPAL · 9 datasets, win/tie/loss | accuracy: 36/9/0 win/tie/loss | Helped |
| reduce initial labeled ratio from 20% to 1%: SPAL advantage over ASPL | SPAL · thyroid | accuracy: 1%: 0.879 vs 0.783; 20%: 0.935 vs 0.938 | Helped |
| SPAL vs ASPL at 20% initial labeled | SPAL · thyroid, 20% initial labels | accuracy: 0.935 vs 0.938 | No clear effect |
Closest papers by meaning
- Navigating the Pitfalls of Active Learning Evaluation: A Systematic Framework for Meaningful Performance Assessment (NeurIPS 2023 · similarity 0.87)
- Learning Active Learning from Data (NeurIPS 2017 · similarity 0.87)
- Search Improves Label for Active Learning (NeurIPS 2016 · similarity 0.87)
- Batch Active Learning at Scale (NeurIPS 2021 · similarity 0.87)
- Re-Active Learning: Active Learning with Relabeling (AAAI 2016 · similarity 0.86)
- Instance-wise Supervision-level Optimization in Active Learning (CVPR 2025 · 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.