An Experimental Study on Fairness-aware Machine Learning for Credit Scoring Problems

TL;DR

This paper conducts a comprehensive experimental evaluation of fairness-aware machine learning models in credit scoring, highlighting their ability to balance accuracy and fairness across financial datasets.

Contribution

It provides the first thorough analysis of fairness-aware models specifically in credit scoring, comparing their performance to traditional models.

Findings

Fairness-aware models improve fairness metrics without significantly sacrificing accuracy.

Certain fairness measures are more effective in credit scoring contexts.

The study offers insights into selecting appropriate fairness measures for financial applications.

Abstract

The digitalization of credit scoring has become essential for financial institutions and commercial banks, especially in the era of digital transformation. Machine learning techniques are commonly used to evaluate customers' creditworthiness. However, the predicted outcomes of machine learning models can be biased toward protected attributes, such as race or gender. Numerous fairness-aware machine learning models and fairness measures have been proposed. Nevertheless, their performance in the context of credit scoring has not been thoroughly investigated. In this paper, we present a comprehensive experimental study of fairness-aware machine learning in credit scoring. The study explores key aspects of credit scoring, including financial datasets, predictive models, and fairness measures. We also provide a detailed evaluation of fairness-aware predictive models and fairness measures on widely used financial datasets. The experimental results show that fairness-aware models achieve a better balance between predictive accuracy and fairness compared to traditional classification models.