Research map / Label-efficient and robust learning / Noisy labels
Active Heteroscedastic Regression
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 weighted least squares instead of ordinary least squares, noise model known | Passive regression · Heteroscedastic linear regression, Gaussian x (theory, simulations) | estimation error bound: ||f*||^2 d/n -> ||f*||^2 (1/n + d^2/n^2) | Helped |
| use active selection of low-noise instances instead of passive sampling, noise model known | Weighted least squares · Heteroscedastic linear regression, Gaussian x (theory, simulations) | estimation error bound: ||f*||^2 (1/n + d^2/n^2) -> ||f*||^2 (1/n + d^2/n^4) | Helped |
| estimate noise model f* then use weighted least squares, instead of ordinary least squares | Passive regression · Heteroscedastic linear regression, Gaussian x (theory, simulations) | estimation error bound: ||f*||^2 d/n -> ||f*||^2 d ln^4 n / n | No clear effect |
| add active selection of low-noise instances with estimated noise model, instead of passive weighted | Two-stage active regression · Heteroscedastic linear regression, Gaussian x (theory, simulations) | estimation error bound: ||f*||^2 d/n -> ||f*||^2 (1/n + d^2/n^2) | Helped |
| use estimated weights with offset gamma=0 instead of offset sqrt(d/n) (ablation) | Weighted least squares with estimated f · Heteroscedastic linear regression, Gaussian x (theory, simulations) | estimation error: error scales as O(d^3/n) | Hurt |
| more unlabeled data m beyond n^2 when noise model estimated | Active regression with estimated noise · Heteroscedastic linear regression, Gaussian x (theory, simulations) | estimation error bound | No clear effect |
| use active learning with noise estimation instead of ordinary least squares | Active regression (Alg 4) · WINE QUALITY | test RMSE | Helped |
| use active learning with noise estimation instead of ordinary least squares | Active regression (Alg 4) · MSD | test RMSE | Helped |
| use weighted least squares with estimated weights instead of ordinary least squares | Passive regression (Alg 3) · WINE QUALITY | test RMSE | Helped |
Closest papers by meaning
- Label Noise: Ignorance Is Bliss (NeurIPS 2024 · similarity 0.85)
- Universal Rates for Active Learning (NeurIPS 2024 · similarity 0.85)
- Semi-Supervised Regression with Heteroscedastic Pseudo-Labels (NeurIPS 2025 · similarity 0.85)
- Towards Self-Supervised Covariance Estimation in Deep Heteroscedastic Regression (ICLR 2025 · similarity 0.85)
- Prediction-Powered Active Testing (NeurIPS 2026 · similarity 0.84)
- Uncertainty Herding: One Active Learning Method for All Label Budgets (ICLR 2025 · 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.