Notes on the H-measure of classifier performance

TL;DR

This paper discusses the H-measure, a flexible classifier performance metric that accounts for application context without fixed misclassification costs, clarifying its interpretation, properties, and relation to other measures.

Contribution

It provides detailed answers to user queries about the H-measure, enhancing understanding of its interpretation, properties, and connections to existing metrics.

Findings

Clarifies the interpretation of the H-measure

Examines the choice of weighting functions

Discusses the measure's propriety and coherence

Abstract

The H-measure is a classifier performance measure which takes into account the context of application without requiring a rigid value of relative misclassification costs to be set. Since its introduction in 2009 it has become widely adopted. This paper answers various queries which users have raised since its introduction, including questions about its interpretation, the choice of a weighting function, whether it is strictly proper, and its coherence, and relates the measure to other work.