Monte Carlo confidence intervals

Paper Title:

Fast Exact/Conservative Monte Carlo Confidence Intervals

Published on:

8 May 2024

Primary Category:

Computation

Paper Authors:

Amanda K. Glazer,

Philip B. Stark

Bullets

Key Details

•

Presents efficient method for Monte Carlo confidence intervals

•

Uses single simulation set for all parameter values

•

Works for exact/conservative Monte Carlo tests

•

Finds intervals quickly when p-value is quasiconcave

•

Additional savings for one-sample and two-sample problems

Explore the topics in this paper

computation reduction

confidence intervals

hypothesis testing

monte carlo methods

python implementation

AI generated summary

Monte Carlo confidence intervals

This paper presents an efficient method to construct exact or conservative confidence intervals from Monte Carlo hypothesis tests. It uses a single set of simulations to test all parameter values, reducing computation. For real-valued parameters, intervals can be found quickly when the p-value is quasiconcave. Additional savings are possible for common statistics in one-sample and two-sample problems. An open-source Python implementation is provided.