Research map / Probabilistic and scientific modeling / Variational and generative sampling
Sequential Predictive Two-Sample and Independence Testing
Aleksandr Podkopaev, Aaditya Ramdas
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 sequential predictive classification instead of the sequential kernelized two-sample test | KDEF · KDEF | observations required to reach perfect power | Improved |
| Use sequential predictive independence testing instead of SKIT | MNIST same-digit dependence · MNIST | data required to achieve perfect power | Improved |
| Use predictive independence testing instead of SKIT | Spherical model · Spherical model | power | Hurt |
| Use predictive independence testing instead of SKIT | HTDD model · HTDD model | power | Improved |
| Use predictive independence testing instead of SKIT | Localized-dependence synthetic data · Localized-dependence synthetic data | power | Improved |
| Use Seq-C-IT instead of Seq-R-IT | Gaussian linear model · Gaussian linear model | power | Improved |
| Use SKIT instead of predictive tests | Gaussian linear model · Gaussian linear model | power | Improved |
| Use adaptive betting fractions rather than a fixed likelihood-ratio-style bet | Learned-predictor sequential two-sample testing · Gaussian two-sample example | test power | Improved |
| Use fixed likelihood-ratio betting rather than adaptive betting fractions | Bayes-optimal predictor sequential two-sample testing · Gaussian two-sample example | test power | Improved |
| Use direct margin maximization instead of minimizing cross-entropy or squared loss | Predictive two-sample testing · Two-sample testing | test power | Hurt |
| Use a one-observation minibatch for the unbalanced-class test | Sequential two-sample testing with unknown class proportions · Unbalanced-class two-sample testing | wealth growth | Hurt |
Closest papers by meaning
- Sequential Kernel Goodness-of-fit Testing (ICML 2024 · similarity 0.90)
- Peeking with PEAK: Sequential, Nonparametric Composite Hypothesis Tests for Means of Multiple Data Streams (ICML 2024 · similarity 0.88)
- DUAL: Learning Diverse Kernels for Aggregated Two-sample and Independence Testing (NeurIPS 2025 · similarity 0.87)
- Optimal Algorithms for Augmented Testing of Discrete Distributions (NeurIPS 2024 · similarity 0.86)
- On the Tightness and Computational Tractability of Higher-Dimensional Confidence Sequences (NeurIPS 2026 · similarity 0.85)
- Protected Test-Time Adaptation via Online Entropy Matching: A Betting Approach (NeurIPS 2024 · 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.
