Research map / Learning theory and algorithms / Sample complexity
Robust Estimation of Derivatives Using Locally Weighted Least Absolute Deviation Regression
WenWu Wang, Ping Yu, Lu Lin, Tiejun Tong
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 |
|---|---|---|---|
| replace symmetric difference with random difference (RLowLAD) (ablation) | LowLAD · first-order derivative, asymptotic variance | asymptotic variance: 75/(16 g(0)^2) -> 75/(24 g(0)^2) | Improved |
| use LowLAD instead of LowLSR under mild normal errors | LowLAD · normal errors N(0,0.1^2), m0 | variance ratio: 0.64 | Hurt |
| use RLowLAD instead of LowLSR under normal errors | RLowLAD · normal errors N(0,0.1^2), m0 | variance ratio: 0.95 | No clear effect |
| use RLowLAD instead of LowLSR under heavy contamination | RLowLAD · normal N(0,1) vs 0.9N(0,1)+0.1N(0,10^2) | variance ratio LowLSR/RLowLAD: 0.95 -> 7.28 | Improved |
| use RLowLAD instead of LAD under bimodal errors | RLowLAD · 0.5N(-3,1)+0.5N(3,1) | variance ratio LAD/RLowLAD: 2.46 × 107 | Improved |
| use RLowLAD instead of LAD under Laplace errors | RLowLAD · Laplace | variance ratio LAD/RLowLAD: 0.75 | Hurt |
| use RLowLAD instead of LowLSR under light-to-bounded distributions | RLowLAD · error distributions with mean 0 variance 1 | ARE: 0.86; 0.90 | Mixed |
| use asymmetric LowLAD instead of one-side LowLAD at the boundary | As-LowLAD · boundary i close to k | variance: one-eighth | Improved |
| use LowLAD instead of LowLSR in contaminated simulations | LowLAD · sigma0 = 10 sigma contamination | AMAE | Improved |
| use LowLAD instead of LowLSR in light contamination | LowLAD · sigma0 = 2 sigma contamination | AMAE | Hurt |
Closest papers by meaning
- Optimal Estimation of Derivatives in Nonparametric Regression (JMLR 2016 · similarity 0.88)
- On the Estimation of Derivatives Using Plug-in Kernel Ridge Regression Estimators (JMLR 2023 · similarity 0.84)
- Smoothed Nonparametric Derivative Estimation using Weighted Difference Quotients (JMLR 2020 · similarity 0.84)
- Distributed High-Dimensional Quantile Regression: Estimation Efficiency and Support Recovery (ICML 2024 · similarity 0.82)
- Relaxed Quantile Regression: Prediction Intervals for Asymmetric Noise (ICML 2024 · similarity 0.80)
- Nonparametric Instrumental Variable Regression through Stochastic Approximate Gradients (NeurIPS 2024 · similarity 0.79)
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.
