Research map / Bandits and online learning / Online convex optimization and games
Online Active Linear Regression via Thresholding
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 |
|---|---|---|---|
| Select observations by thresholding their whitened, weighted norm instead of random sampling | Online linear regression with OLS · Physicochemical Properties of Protein Tertiary Structure | test squared prediction error | Helped |
| Select observations by thresholding their whitened, weighted norm instead of random sampling | Online linear regression with OLS · Bike Sharing | MSE | Helped |
| Select observations by thresholding their whitened, weighted norm instead of random sampling | Online linear regression with OLS · YearPredictionMSD | MSE | Helped |
| Update the threshold and weights adaptively after each observation instead of fixing them in advance (ablation) | Thresholding Algorithm · Synthetic Gaussian linear model | MSE | Helped |
| Use optimal weights rather than uniform weights (ablation) | Thresholding Algorithm · Correlated Laplace marginals | MSE | Helped |
| Use approximate uniform weights instead of optimal weights (ablation) | Thresholding Algorithm · Correlated Laplace marginals | MSE | Hurt |
| Use thresholding rather than random sampling | Linear regression under increasing model nonlinearity · Synthetic nonlinear data | MSE | Mixed |
| Use thresholding rather than random sampling | Ridge regression · Synthetic white Gaussian linear model | MSE: less than 30% | Helped |
| Use thresholding rather than random sampling | Lasso regression · Sparse synthetic model, varying effective dimension | MSE | Helped |
| Use thresholding rather than random sampling | Lasso regression · Sparse synthetic model, increasing ambient dimension | MSE | Helped |
| Use a two-stage procedure that first estimates the support and then thresholds observations in the selected subspace | Sparse Thresholding Algorithm · High-dimensional sparse linear regression | MSE: at least a log d factor | Helped |
| Fit the final sparse-model estimate using all k observations rather than only second-stage observations (ablation) | Sparse Thresholding Algorithm · High-dimensional sparse linear regression | MSE | Helped |
Closest papers by meaning
- Online Active Regression (ICML 2022 · similarity 0.90)
- Improved Leverage Score Sampling for Constrained Active Linear Regression (NeurIPS 2026 · similarity 0.88)
- Improved Active Learning via Dependent Leverage Score Sampling (ICLR 2024 · similarity 0.85)
- Online Linear Regression in Dynamic Environments via Discounting (ICML 2024 · similarity 0.85)
- Near-optimal Active Regression of Single-Index Models (ICLR 2025 · similarity 0.85)
- Robust Sampling for Active Statistical Inference (NeurIPS 2025 · similarity 0.85)
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.