Dual Confidence Regions: A Simple Introduction - DataScienceCentral.com

This tutorial explains how to build confidence regions (the 2D version of a confidence interval) using as little statistical theory as possible. I also avoid the traditional terminology and notation such as α, Z1-α, critical value, confidence level, significance level and so on. These can be confusing to beginners and professionals alike. Instead, I use simulations and two keywords only: confidence region, and confidence level. The purpose is to explain the concept using a framework that will appeal to machine learning professionals, software engineers and non-statisticians.