Research map / Learning theory and algorithms / Kernels and regression theory
Consistent Structured Prediction with Max-Min Margin Markov Networks
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 M⁴N instead of M³N | multi-class classification · segment | average test loss: 6.64% -> 6.09% | Helped |
| Use M⁴N instead of the strongest baseline | multi-class classification · iris | average test loss: 3.08% -> 3.33% | Hurt |
| Use M⁴N instead of the strongest baseline | multi-class classification · wine (MC) | average test loss: 2.14% -> 2.35% | Hurt |
| Use M⁴N instead of M³N | multi-class classification · vehicle (MC) | average test loss: 24.6% -> 24% | Helped |
| Use M⁴N instead of the strongest baseline | multi-class classification · satimage | average test loss: 11.5% -> 11.9% | Hurt |
| Use M⁴N instead of the strongest baseline | multi-class classification · letter | average test loss: 13.2% -> 13.5% | Hurt |
| Use M⁴N instead of M³N | multi-class classification · mfeat | average test loss: 3.94% -> 3.96% | Hurt |
| Use M⁴N instead of M³N | ordinal regression · stocks | average test loss: 0.167 -> 0.160 | Helped |
| Use M⁴N instead of M³N | ordinal regression · machine | average test loss: 0.634 -> 0.628 | Helped |
| Use M⁴N instead of M³N | ordinal regression · abalone | average test loss: 0.520 -> 0.520 | No clear effect |
| Use M⁴N instead of M³N | sequence prediction · OCR | average test loss: 16.2% -> 16.2% | No clear effect |
| Use M⁴N instead of M³N | ranking · wine (ranking) | average test loss: 5.06% -> 4.34% | Helped |
Closest papers by meaning
- Max-Margin Works while Large Margin Fails: Generalization without Uniform Convergence (ICLR 2023 · similarity 0.87)
- Learning latent variable structured prediction models with Gaussian perturbations (NeurIPS 2018 · similarity 0.87)
- A Smoother Way to Train Structured Prediction Models (NeurIPS 2018 · similarity 0.87)
- Structured Prediction Theory Based on Factor Graph Complexity (NeurIPS 2016 · similarity 0.87)
- The Performance Analysis of Generalized Margin Maximizers on Separable Data (ICML 2020 · similarity 0.87)
- Multi-Class Support Vector Machine with Maximizing Minimum Margin (AAAI 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.