Time for a change: a tutorial for comparing multiple classifiers through Bayesian analysis

TL;DR

This paper advocates replacing traditional null hypothesis significance testing with Bayesian analysis for comparing multiple classifiers, emphasizing more sound and reliable statistical methods in machine learning research.

Contribution

It introduces Bayesian methods as superior alternatives for classifier comparison, criticizing NHST and providing practical guidance for adoption.

Findings

Bayesian analysis offers more reliable classifier comparisons.

NHST has significant fallacies that undermine scientific validity.

The paper provides a tutorial for implementing Bayesian methods.

Abstract

The machine learning community adopted the use of null hypothesis significance testing (NHST) in order to ensure the statistical validity of results. Many scientific fields however realized the shortcomings of frequentist reasoning and in the most radical cases even banned its use in publications. We should do the same: just as we have embraced the Bayesian paradigm in the development of new machine learning methods, so we should also use it in the analysis of our own results. We argue for abandonment of NHST by exposing its fallacies and, more importantly, offer better - more sound and useful - alternatives for it.