Research map / Learning theory and algorithms / Prediction and decision losses
Localized Structured Prediction
Carlo Ciliberto, Francis Bach, Alessandro Rudi
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 Local-Δ with part-based locality instead of Δ-Global | Localized Structured Prediction · FVC04 DB1_B fingerprint ridge orientation | average pixel-wise geodesic error | Improved |
| Use Local-LS instead of Δ-Global | Localized Structured Prediction · FVC04 DB1_B fingerprint ridge orientation | average pixel-wise geodesic error | Improved |
| Replace the Δ loss with least-squares while retaining locality (ablation) | Localized Structured Prediction · FVC04 DB1_B fingerprint ridge orientation | average pixel-wise geodesic error | Hurt |
| Use Local-LS instead of Global-LS | Linear multitask regression · Synthetic within-locality simulation | mean square error | Improved |
| Use Local-LS instead of IndependentParts-LS, sharing the local relation across parts | Linear multitask regression · Synthetic within-locality simulation | mean square error | Improved |
| Increase within-locality decay γ (ablation) | Localized Structured Prediction (Local-LS) · Synthetic within-locality simulation | mean square error | Improved |
| Increase the number of parts |P| (ablation) | Localized Structured Prediction (Local-LS) · Synthetic within-locality simulation | mean square error | Improved |
| Use locality-aware learning when input parts are highly correlated (γ = 0) (ablation) | Localized Structured Prediction (Local-LS) · Synthetic within-locality simulation | mean square error | No clear effect |
Closest papers by meaning
- A General Framework for Consistent Structured Prediction with Implicit Loss Embeddings (JMLR 2020 · similarity 0.86)
- A Smoother Way to Train Structured Prediction Models (NeurIPS 2018 · similarity 0.86)
- Learning latent variable structured prediction models with Gaussian perturbations (NeurIPS 2018 · similarity 0.85)
- On Certified Generalization in Structured Prediction (NeurIPS 2023 · similarity 0.85)
- Mitigating the Effect of Incidental Correlations on Part-based Learning (NeurIPS 2023 · similarity 0.85)
- Role of Locality and Weight Sharing in Image-Based Tasks: A Sample Complexity Separation between CNNs, LCNs, and FCNs (ICLR 2024 · 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.
