Addressing Fairness, Bias and Class Imbalance in Machine Learning: the FBI-loss

05/13/2021
by   Elisa Ferrari, et al.
0

Resilience to class imbalance and confounding biases, together with the assurance of fairness guarantees are highly desirable properties of autonomous decision-making systems with real-life impact. Many different targeted solutions have been proposed to address separately these three problems, however a unifying perspective seems to be missing. With this work, we provide a general formalization, showing that they are different expressions of unbalance. Following this intuition, we formulate a unified loss correction to address issues related to Fairness, Biases and Imbalances (FBI-loss). The correction capabilities of the proposed approach are assessed on three real-world benchmarks, each associated to one of the issues under consideration, and on a family of synthetic data in order to better investigate the effectiveness of our loss on tasks with different complexities. The empirical results highlight that the flexible formulation of the FBI-loss leads also to competitive performances with respect to literature solutions specialised for the single problems.

READ FULL TEXT

page 14

page 15

research
09/13/2022

Investigating Bias with a Synthetic Data Generator: Empirical Evidence and Philosophical Interpretation

Machine learning applications are becoming increasingly pervasive in our...
research
07/02/2018

A Broader View on Bias in Automated Decision-Making: Reflecting on Epistemology and Dynamics

Machine learning (ML) is increasingly deployed in real world contexts, s...
research
09/21/2021

Fairness-aware Class Imbalanced Learning

Class imbalance is a common challenge in many NLP tasks, and has clear c...
research
08/31/2019

Imbalance Problems in Object Detection: A Review

In this paper, we present a comprehensive review of the imbalance proble...
research
07/30/2021

Foundations of data imbalance and solutions for a data democracy

Dealing with imbalanced data is a prevalent problem while performing cla...
research
05/23/2023

Fair Oversampling Technique using Heterogeneous Clusters

Class imbalance and group (e.g., race, gender, and age) imbalance are ac...
research
07/02/2020

Tilted Empirical Risk Minimization

Empirical risk minimization (ERM) is typically designed to perform well ...

Please sign up or login with your details

Forgot password? Click here to reset