Research map / Optimization / Optimizers
Online Bootstrap Confidence Intervals for the Stochastic Gradient Descent Estimator
Yixin Fang, Jinfeng Xu, Lei Yang
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 random-weighting bootstrap (RW-σ) in place of recursive plug-in for SE | SGD for linear regression · Linear regression (10000,10,6,0.1) | coverage of 95% CI: 0.953 -> 0.956 | No clear effect |
| use RW-σ instead of RPI | SGD for logistic regression · Logistic regression (10000,10,6,0.2) | coverage of 95% CI: 0.955 -> 0.959 | No clear effect |
| apply random weighting where plug-in is not applicable | LAD regression SGD · LAD Setting 3 (10000,10,6,0.1) | coverage of 95% CI: 0.965 and 0.969, SE 0.029 vs empirical 0.026 | Improved |
| apply random weighting with outlier-contaminated data | LAD regression SGD · LAD Setting 4 (10000,10,6,0.1) | coverage of 95% CI: 0.954 and 0.958 | No clear effect |
| increase sample size from N=10000 to N=20000 (p=10 to 20) | RW-σ standard error, linear regression · Linear regression | standard error: 0.012 -> 0.008 | Improved |
Closest papers by meaning
- Cheap Bootstrap for Fast Uncertainty Quantification of Stochastic Gradient Descent (JMLR 2026 · similarity 0.92)
- HiGrad: Uncertainty Quantification for Online Learning and Stochastic Approximation (JMLR 2023 · similarity 0.87)
- On the Utility of Equal Batch Sizes for Inference in Stochastic Gradient Descent (JMLR 2025 · similarity 0.87)
- Fast and Robust Online Inference with Stochastic Gradient Descent via Random Scaling (AAAI 2022 · similarity 0.87)
- Risk Bounds of Accelerated SGD for Overparameterized Linear Regression (ICLR 2024 · similarity 0.85)
- Learning Curves of Stochastic Gradient Descent in Kernel Regression (ICML 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.
