Incertitudes et mesures

TL;DR

This educational guide explains how to statistically handle measurement uncertainties, covering classical methods and introducing Bayesian inference with practical applications like failure analysis and noise reduction.

Contribution

It provides a detailed explanation of measurement uncertainty treatment, including classical and Bayesian methods, with specific application cases.

Findings

Clarifies conditions for applying statistical practices

Introduces Bayesian inference in measurement analysis

Demonstrates practical applications in engineering contexts

Abstract

Educational guide focused on the statistical treatment of measurement uncertainties. The conditions of application of current practices are detailed and precised: mean values, central limit theorem, linear regression. The last two chapters are devoted to an introduction to the Bayesian inference and a series of application cases: machine failure date, elimination of a background noise, linear adjustment with elimination of outliers.