Research map / Learning theory and algorithms / Prediction and decision losses
Boosting with Abstention
Corinna Cortes, Giulia DeSalvo, Mehryar Mohri
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Learn the abstention function jointly with the predictor (BA) instead of the confidence-based two-step TSB | BA algorithm with L_SB · australian, cod, skin, banknote, haberman, pima | average rejection loss | Improved |
| Use BA with abstention stumps instead of the DHL algorithm (polynomial kernels) | BA algorithm with L_SB · three datasets in the second row of Figure 5 | average rejection loss | Improved |
| Use confidence-based TSB instead of DHL | TSB algorithm · australian, cod, skin, banknote, haberman, pima | average rejection loss | Mixed |
| Scale to CIFAR-10 (boat and horse images, 10,000 instances): run BA instead of DHL | BA algorithm · CIFAR-10 boat vs horse | training feasibility | Improved |
Closest papers by meaning
- Revisiting Agnostic Boosting (NeurIPS 2025 · similarity 0.86)
- Margins are Insufficient for Explaining Gradient Boosting (NeurIPS 2020 · similarity 0.86)
- Efficient Active Learning with Abstention (NeurIPS 2022 · similarity 0.86)
- Margin-Based Generalization Lower Bounds for Boosted Classifiers (NeurIPS 2019 · similarity 0.85)
- Sample-Efficient Agnostic Boosting (NeurIPS 2024 · similarity 0.85)
- The Many Faces of Optimal Weak-to-Strong Learning (NeurIPS 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.
