English

Game-theoretic statistics and safe anytime-valid inference

Statistics Theory 2023-06-21 v2 Computer Science and Game Theory Information Theory math.IT Methodology Statistics Theory

Abstract

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 stopping times, accommodating continuous monitoring and analysis of accumulating data and optional stopping or continuation for any reason. These measures crucially rely on test martingales, which are nonnegative martingales starting at one. Since a test martingale is the wealth process of a player in a betting game, SAVI centrally employs game-theoretic intuition, language and mathematics. We summarize the SAVI goals and philosophy, and report recent advances in testing composite hypotheses and estimating functionals in nonparametric settings.

Cite

@article{arxiv.2210.01948,
  title  = {Game-theoretic statistics and safe anytime-valid inference},
  author = {Aaditya Ramdas and Peter Grünwald and Vladimir Vovk and Glenn Shafer},
  journal= {arXiv preprint arXiv:2210.01948},
  year   = {2023}
}

Comments

25 pages. Under review. ArXiv does not compile/space some references properly

R2 v1 2026-06-28T02:49:06.200Z