Game-theoretic statistics and safe anytime-valid inference

A Ramdas, P Grünwald, V Vovk, G Shafer - Statistical Science, 2023 - projecteuclid.org
Safe anytime-valid inference (SAVI) provides measures of statistical evidence and certainty—
e-processes for testing and confidence sequences for estimation—that remain valid at all …

Estimating means of bounded random variables by betting

I Waudby-Smith, A Ramdas - Journal of the Royal Statistical …, 2024 - academic.oup.com
We derive confidence intervals (CIs) and confidence sequences (CSs) for the classical
problem of estimating a bounded mean. Our approach generalizes and improves on the …

E-values: Calibration, combination and applications

V Vovk, R Wang - The Annals of Statistics, 2021 - projecteuclid.org
E-values: Calibration, combination and applications Page 1 The Annals of Statistics 2021, Vol.
49, No. 3, 1736–1754 https://doi.org/10.1214/20-AOS2020 © Institute of Mathematical Statistics …

Admissible anytime-valid sequential inference must rely on nonnegative martingales

A Ramdas, J Ruf, M Larsson, W Koolen - arxiv preprint arxiv:2009.03167, 2020 - arxiv.org
Confidence sequences, anytime p-values (called p-processes in this paper), and e-
processes all enable sequential inference for composite and nonparametric classes of …

Likelihood ratio confidence sets for sequential decision making

N Emmenegger, M Mutny… - Advances in Neural …, 2024 - proceedings.neurips.cc
Certifiable, adaptive uncertainty estimates for unknown quantities are an essential
ingredient of sequential decision-making algorithms. Standard approaches rely on problem …

False discovery rate control with e-values

R Wang, A Ramdas - Journal of the Royal Statistical Society …, 2022 - academic.oup.com
E-values have gained attention as potential alternatives to p-values as measures of
uncertainty, significance and evidence. In brief, e-values are realized by random variables …

Derandomised knockoffs: leveraging e-values for false discovery rate control

Z Ren, RF Barber - Journal of the Royal Statistical Society Series …, 2024 - academic.oup.com
Abstract Model-X knockoffs is a flexible wrapper method for high-dimensional regression
algorithms, which provides guaranteed control of the false discovery rate (FDR). Due to the …

Testing exchangeability: Fork-convexity, supermartingales and e-processes

A Ramdas, J Ruf, M Larsson, WM Koolen - International Journal of …, 2022 - Elsevier
Suppose we observe an infinite series of coin flips X 1, X 2,…, and wish to sequentially test
the null that these binary random variables are exchangeable. Nonnegative …

Improved regret bounds of (multinomial) logistic bandits via regret-to-confidence-set conversion

J Lee, SY Yun, KS Jun - International Conference on …, 2024 - proceedings.mlr.press
Logistic bandit is a ubiquitous framework of modeling users' choices, eg, click vs. no click for
advertisement recommender system. We observe that the prior works overlook or neglect …

A Discussion of" A Note on Universal Inference" by Tse and Davison.

A Spector, E Candès, L Lei - Stat, 2023 - search.ebscohost.com
The article discusses the concept of universal inference (UI) and its limitations in problems
with high-dimensional nuisance parameters. It introduces a modification called the quasi …