Machine Learning Fairness for Depression Detection using EEG Data

TL;DR

This study evaluates fairness in machine learning models for depression detection using EEG data, revealing existing biases and assessing mitigation strategies across various deep learning architectures and datasets.

Contribution

First comprehensive analysis of fairness in EEG-based depression detection, comparing multiple bias mitigation methods across different neural network models and datasets.

Findings

Bias exists in EEG datasets and algorithms for depression detection.

Bias mitigation methods can reduce bias at different stages and levels.

Different fairness measures respond variably to mitigation strategies.

Abstract

This paper presents the very first attempt to evaluate machine learning fairness for depression detection using electroencephalogram (EEG) data. We conduct experiments using different deep learning architectures such as Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM) networks, and Gated Recurrent Unit (GRU) networks across three EEG datasets: Mumtaz, MODMA and Rest. We employ five different bias mitigation strategies at the pre-, in- and post-processing stages and evaluate their effectiveness. Our experimental results show that bias exists in existing EEG datasets and algorithms for depression detection, and different bias mitigation methods address bias at different levels across different fairness measures.