Notes on the H-measure of classifier performance
Authors:
Preprint
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.Keywords
All Related Versions
This publication has 12 references indexed in Scilit:
- Performance measures in evaluating machine learning based bioinformatics predictors for classificationsQuantitative Biology, 2016
- A better Beta for the H measure of classification performancePattern Recognition Letters, 2014
- When is the area under the receiver operating characteristic curve an appropriate measure of classifier performance?Pattern Recognition Letters, 2013
- Evaluating diagnostic tests: The area under the ROC curve and the balance of errorsStatistics in Medicine, 2010
- ROC Curves for Continuous Data2009
- Strictly Proper Scoring Rules, Prediction, and EstimationJournal of the American Statistical Association, 2007
- The geometry of proper scoring rulesAnnals of the Institute of Statistical Mathematics, 2006
- The Area under the ROC Curve and Its CompetitorsMedical Decision Making, 1991
- A General Method for Comparing Probability AssessorsThe Annals of Statistics, 1989
- Admissible Probability Measurement ProceduresPsychometrika, 1966