Research map / Learning theory and algorithms / Kernels and regression theory
On Margins and Generalisation for Voting Classifiers
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 the Dirichlet margin bound instead of existing margin bounds | Voting classifiers · Classification datasets from UCI, LIBSVM, and Zalando | generalization bound | Helped |
| Use the Dirichlet margin bound instead of the PAC-Bayes bound used as the optimization objective | Outputs of PAC-Bayes algorithms · Binary and multiclass classification datasets | generalization bound | Helped |
| Use the optimization-friendly expected-margin-loss variant as the training objective | Voting classifiers with random-forest voters · Classification datasets | test error | No clear effect |
| Expand the search space for K | Dirichlet margin bound with decision stumps · ADULT | generalization bound: 0.36 ± 0.10 | Helped |
| Optimize the Dirichlet margin bound directly rather than apply it algorithm-free to weights from another algorithm | Voting classifiers with decision-stump voters · Classification datasets | generalization bound | Mixed |
Closest papers by meaning
- Tight Margin-Based Generalization Bounds for Voting Classifiers over Finite Hypothesis Sets (ICML 2026 · similarity 0.93)
- Learning Stochastic Majority Votes by Minimizing a PAC-Bayes Generalization Bound (NeurIPS 2021 · similarity 0.92)
- How many classifiers do we need? (NeurIPS 2024 · similarity 0.86)
- Exact power indices for plurality-voting ensembles (NeurIPS 2026 · similarity 0.86)
- On Rate-Optimal Partitioning Classification from Observable and from Privatised Data (NeurIPS 2026 · similarity 0.85)
- Controlling Multiple Errors Simultaneously with a PAC-Bayes Bound (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.