Why ROC-AUC Is Misleading for Highly Imbalanced Data: In-Depth Evaluation of MCC, F2-score, H-measure, and AUC-based Metrics across Diverse Classifiers

Preprints.org 2026 Preprint Open access

Read full text
Authors: Mehdi Imani, Majid Joudaki, Ayoub Bagheri, Hamid Reza Arabnia
DOI: 10.20944/preprints202510.0958.v2
Cited by: 3

Abstract

This study re-evaluates ROC-AUC for binary classification under severe class imbalance (

Fields of Study & Subjects

Imbalanced Data Classification Techniques Explainable Artificial Intelligence (XAI) Financial Distress and Bankruptcy Prediction

Publication details

Journal / source
Preprints.org
Publisher
MDPI AG
Publication date
13 January 2026
Document type
Preprint