Fairness in Data Science: application of methodologies for the mitigation of social biases in a Machine Learning model for credit granting

Vol 56, 2024 - 309823
Complete Articles (CA)
Favorite this paper
How to cite this paper?
Abstract

The discussion about the impacts of social biases on the fairness of Machine Learning models has been increasing its relevance nowadays with the greater use of automated software where human interference is not necessary. This article focuses on selecting methodologies that attempt to remove social bias and applying them in the context of credit concession utilizing a Brazilian database. The methodologies employed utilize different rationales for bias removal, including resampling to rebalance the dataset, modifying the target variable to achieve greater balance, and removing sensitive variable information from the predictor variables. All methodologies managed to remove bias from the dataset to some extent. However, this removal is accompanied by a loss of model performance.

Share your ideas or questions with the authors!

Did you know that the greatest stimulus in scientific and cultural development is curiosity? Leave your questions or suggestions to the author!

Sign in to interact

Have a question or suggestion? Share your feedback with the authors!

Institutions
  • 1 Universidade Federal de Minas Gerais
  • 2 UFMG
Track
  • 9. EST&MP – Statistics and Probabilistic Models
Keywords
Social Biases
Fairness
Machine Learning