Research map / Learning theory and algorithms / Kernels and regression theory
Sufficient reductions in regression with mixed predictors
Efstathia Bura, Liliana Forzani, Rodrigo García Arancibia, Pamela Llop, Diego Tomassi
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 SDR reduction instead of full unreduced logistic regression | Logistic regression · Krzanowski set 1 | misclassification rate: 0.375 -> 0.250 | Improved |
| use sub-optimal SDR instead of full model | Logistic regression · Krzanowski set 2 | misclassification rate: 0.258 -> 0.204 | Improved |
| use sub-optimal SDR instead of Krzanowski location model | Location model · Krzanowski set 4 | misclassification rate: 0.328 -> 0.290 | Improved |
| sub-optimal vs optimal SDR (ablation) | SDR · Krzanowski set 1 | misclassification rate: 0.250 -> 0.300 | Hurt |
| sub-optimal vs optimal SDR (ablation) | SDR · Krzanowski set 3 | misclassification rate: 0.161 -> 0.145 | Improved |
| replace PCA with PCAmix on mixed predictors | PCA · Krzanowski set 1 | misclassification rate: 0.325 -> 0.425 | Hurt |
| SDR Optimal vs PCA reduction | PCA · Krzanowski set 3 | misclassification rate: 0.484 -> 0.161 | Improved |
| sub-optimal SDR vs full model in AUC | Full logistic · Krzanowski set 4 | AUC: 0.738 -> 0.785 | Improved |
| replace PCA by supervised PFC without country effect | PCA · Governance, linear | MSPE: 0.319 -> 0.292 | Improved |
| replace PCA by PFC without country effect | PCA · Governance, kernel | MSPE: 0.189 -> 0.282 | Hurt |
| add country effect via mixed optimal SDR | PFC · Governance, linear | MSPE: 0.292 -> 0.029 | Improved |
| replace PCA with PCAmix on governance with country effect | PCA · Governance, linear | MSPE: 0.319 -> 0.320 | No clear effect |
Closest papers by meaning
- MixLasso: Generalized Mixed Regression via Convex Atomic-Norm Regularization (NeurIPS 2018 · similarity 0.84)
- List-decodable Linear Regression (NeurIPS 2019 · similarity 0.83)
- Balanced Sharpness-Aware Minimization for Imbalanced Regression (ICCV 2025 · similarity 0.83)
- Sparse Mixed Linear Regression with Guarantees: Taming an Intractable Problem with Invex Relaxation (ICML 2022 · similarity 0.83)
- Agnostic Learning of Mixed Linear Regressions with EM and AM Algorithms (ICML 2024 · similarity 0.83)
- Regression with Multi-Expert Deferral (ICML 2024 · similarity 0.81)
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.
