English

Setting Confidence Belts

High Energy Physics - Experiment 2009-10-31 v2

Abstract

We propose using a Bayes procedure with uniform improper prior to determine credible belts for the mean of a Poisson distribution in the presence of background and for the continuous problem of measuring a non-negative quantity θ\theta with a normally distributed measurement error. Within the Bayesian framework, these belts are optimal. The credible limits are then examined from a frequentist point of view and found to have good frequentist and conditional frequentist properties.

Keywords

Cite

@article{arxiv.hep-ex/0007048,
  title  = {Setting Confidence Belts},
  author = {Byron P. Roe and Michael B. Woodroofe},
  journal= {arXiv preprint arXiv:hep-ex/0007048},
  year   = {2009}
}

Comments

26 pages, 8 figures, submitted to Phys. Rev. D