Sequential testing for elicitable functionals via supermartingales

DOI10.3150/23-BEJ1634arXiv2204.05680OpenAlexW4391458275WikidataQ128886106 ScholiaQ128886106MaRDI QIDQ6201853FDOQ6201853

Authors: Philippe Casgrain, Martin Larsson, Johanna F. Ziegel 

Publication date: 26 March 2024

Published in: Bernoulli (Search for Journal in Brave)

Abstract: We design sequential tests for a large class of nonparametric null hypotheses based on elicitable and identifiable functionals. Such functionals are defined in terms of scoring functions and identification functions, which are ideal building blocks for constructing nonnegative supermartingales under the null. This in turn yields sequential tests via Ville's inequality. Using regret bounds from Online Convex Optimization, we obtain rigorous guarantees on the asymptotic power of the tests for a wide range of alternative hypotheses. Our results allow for bounded and unbounded data distributions, assuming that a sub-$psi$ tail bound is satisfied.

Full work available at URL: https://arxiv.org/abs/2204.05680

zbMATH Keywords

online optimizationsequential statisticsanytime valid testingelicitable functionalsidentifiable functionals

Mathematics Subject Classification ID

Statistics (62-XX) Information and communication theory, circuits (94-XX)

Cites Work

• On tail probabilities for martingales
• Higher order elicitability and Osband's principle
• Semiparametric efficiency bound in time-series models for conditional quantiles
• Making and evaluating point forecasts
• Time-uniform, nonparametric, nonasymptotic confidence sequences
• Valid sequential inference on probability forecast performance
• Sequential Tests of Statistical Hypotheses
• Test martingales, Bayes factors and \(p\)-values
• E-values: calibration, combination and applications
• Elicitation of Personal Probabilities and Expectations
• Multiple Comparisons
• Online learning and online convex optimization
• A guided tour of Chernoff bounds
• Efficient estimation in dynamic conditional quantile models
• Self-normalized processes: exponential inequalities, moment bounds and iterated logarithm laws.
• Title not available (Why is that?)
• Title not available (Why is that?)
• The collected works of Wassily Hoeffding. Ed. by N. I. Fisher and P. K. Sen
• A survey of algorithms and analysis for adaptive online learning
• Time-uniform Chernoff bounds via nonnegative supermartingales
• Elicitation complexity of statistical properties
• Testing exchangeability: fork-convexity, supermartingales and e-processes
• Memoryless sequences for general losses

This page was built for publication: Sequential testing for elicitable functionals via supermartingales

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6201853)