Research map / Learning theory and algorithms / Sample complexity
Beyond First-order Asymptotics in Sequential Mean Testing
VIKAS DEEP, Shubhada Agrawal
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 practical constant threshold beta=log(1/alpha) instead of theoretical beta=1+log(2(1+n)/alpha) (ablation) | KLinf sequential test · Bernoulli(0.6), m0=0.2, stopping-time CLT | Gaussianity of normalized stopping time | Improved |
| decrease alpha from 1e-4 to 1e-8 (ablation) | KLinf sequential test · Bernoulli(0.6), m0=0.2 | Gaussian fit of normalized stopping time | Improved |
| increase sample size n from 1000 to 2000 (ablation) | empirical KLinf statistic · Beta(3,2) and Bernoulli(0.6), m0=0.7 | closeness to Gaussian limit | Improved |
Closest papers by meaning
- Optimal sequential tests yield log-optimal e-processes (NeurIPS 2026 · similarity 0.87)
- Asymptotic Anytime-Valid Inference for U-statistics (NeurIPS 2026 · similarity 0.85)
- Mean Testing under Truncation beyond Gaussian (NeurIPS 2026 · similarity 0.84)
- Peeking with PEAK: Sequential, Nonparametric Composite Hypothesis Tests for Means of Multiple Data Streams (ICML 2024 · similarity 0.84)
- Optimal Rates for Differentially Private Hypothesis Testing with E-values (NeurIPS 2026 · similarity 0.84)
- Sequential Kernel Goodness-of-fit Testing (ICML 2024 · similarity 0.83)
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.
