Research map / Learning theory and algorithms / Sample complexity
Linear Maximum Margin Classifier for Learning from Uncertain Data
Christos Tzelepis, Vasileios Mezaris, Ioannis Patras
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 |
|---|---|---|---|
| use full Gaussian sample covariance (GSU) instead of ignoring uncertainty | linear SVM · MNIST 1 vs 7 (D5) | testing accuracy: 0.6027 -> 0.6640 | Improved |
| learn in linear subspaces preserving a fraction of variance instead of the original space (ablation) | LSVM-GSU · MNIST 1 vs 7 (D3) | testing accuracy: 0.7216 -> 0.7543 | Improved |
| learn in linear subspaces instead of original space (ablation) | LSVM-GSU · MNIST 1 vs 7 (D2) | testing accuracy: 0.8310 -> 0.8562 | Improved |
| model uncertainty anisotropically (GSU) instead of isotropically | LSVM-iso · MNIST 1 vs 7 (D4) | testing accuracy: 0.6675 -> 0.6974 | Improved |
| learn in original space with GSU vs LSVM-iso on D3 | LSVM-iso · MNIST 1 vs 7 (D3) | testing accuracy: 0.7222 -> 0.7216 | No clear effect |
| use LSVM-GSU instead of LSVM | linear SVM · WDBC | testing accuracy: 95.15% -> 97.14% | Improved |
| use LSVM-GSU instead of LSVM-iso | linear SVM · WDBC | testing accuracy: 96.53% -> 97.14% | Improved |
| use LSVM-GSU instead of LSVM | linear SVM · DEAP arousal | accuracy: 0.626 -> 0.659 | Improved |
| use LSVM-GSU instead of NB | classifier · DEAP arousal | F1: 0.583 -> 0.551 | Hurt |
| use LSVM-GSU with partial variance information | linear SVM · TV news (train CNN, test CNN) | F1: 0.7799 -> 0.9589 | Improved |
| use LSVM-GSU when training on NDTV and testing on TIMES NOW | linear SVM · TV news | F1: 0.7809 -> 0.7493 | Hurt |
| use LSVM-GSU instead of LSVM, config 1 | linear SVM · MED14Test | MAP: 0.1478 -> 0.1806 | Improved |
Closest papers by meaning
- Soft Margin Support Vector Classification as Buffered Probability Minimization (JMLR 2017 · similarity 0.88)
- The Performance Analysis of Generalized Margin Maximizers on Separable Data (ICML 2020 · similarity 0.87)
- Risk Bounds for Over-parameterized Maximum Margin Classification on Sub-Gaussian Mixtures (NeurIPS 2021 · similarity 0.87)
- Multi-Class Support Vector Machine with Maximizing Minimum Margin (AAAI 2024 · similarity 0.86)
- Polynomial-Time Robust Multiclass Linear Classification under Gaussian Marginals (NeurIPS 2026 · similarity 0.84)
- Minimax Optimal Early-Stopped Gradient Descent for Gaussian Mixture Classification (NeurIPS 2026 · similarity 0.83)
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.
