Research map / Learning theory and algorithms / Sample complexity
Ultra-High Dimensional Single-Index Quantile Regression
Yuankun Zhang, Heng Lian, Yan Yu
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 single-index quantile model with SCAD instead of linear quantile model with SCAD | penalized quantile regression · Mouse gene expression PEPCK (tau=0.5) | mean check-loss prediction error: 0.277 -> 0.114 | Improved |
| use single-index quantile model with SCAD instead of linear quantile model with SCAD | penalized quantile regression · Mouse gene expression PEPCK (tau=0.25) | mean check-loss prediction error: 0.158 -> 0.081 | Improved |
| use single-index quantile model with SCAD instead of linear quantile model with SCAD | penalized quantile regression · Mouse gene expression PEPCK (tau=0.75) | mean check-loss prediction error: 0.186 -> 0.084 | Improved |
| use single-index model instead of linear lasso | penalized quantile regression · Mouse gene expression PEPCK (tau=0.5) | model size (selected genes): 60 -> 11 | Improved |
| use SCAD instead of MCP penalty in single-index model (ablation) | single-index quantile regression · Mouse gene expression PEPCK (tau=0.5) | mean check-loss prediction error: 0.141 -> 0.114 | No clear effect |
| use SCAD instead of MCP penalty in single-index model (ablation) | single-index quantile regression · Mouse gene expression PEPCK (tau=0.75) | mean check-loss prediction error: 0.073 -> 0.084 | No clear effect |
| increase covariates p from 50 to 1000 (n=200) | hdsim SCAD · Heteroscedastic simulation (tau=0.5) | MSE (x1e-4): 3.21 -> 3.33 | No clear effect |
| penalized selection versus oracle knowing the true variables | hdsim SCAD · Heteroscedastic simulation (n=100, p=50, tau=0.5) | MSE (x1e-4): 4.88 -> 6.18 | Hurt |
| increase sample size from 100 to 200 | hdsim SCAD · Heteroscedastic simulation (p=50, tau=0.5) | MSE (x1e-4): 6.18 -> 3.21 | Improved |
Closest papers by meaning
- Additive Nonlinear Quantile Regression in Ultra-high Dimension (JMLR 2022 · similarity 0.91)
- Non-parametric Quantile Regression via the K-NN Fused Lasso (JMLR 2021 · similarity 0.88)
- Confidence Intervals and Hypothesis Testing for High-dimensional Quantile Regression: Convolution Smoothing and Debiasing (JMLR 2023 · similarity 0.88)
- High-Dimensional Structured Quantile Regression (ICML 2017 · similarity 0.87)
- Distributed High-Dimensional Quantile Regression: Estimation Efficiency and Support Recovery (ICML 2024 · similarity 0.85)
- Robust group and simultaneous inferences for high-dimensional single index model (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.
