Toward Mitigating Sex Bias in Pilot Trainees' Stress and Fatigue Modeling
Rachel Pfeifer, Sudip Vhaduri, Mark Wilson, Julius Keller
TL;DR
The paper tackles sex bias in stress/fatigue models for pilot trainees, a domain with a markedly skewed gender distribution that can compromise model fairness and safety. It applies two bias-mitigation strategies—threshold optimization with demographic parity and equalized odds constraints—on decision-tree classifiers trained with pilot and non-pilot data, repeating the evaluation across 30 random instantiations. The results show substantial fairness improvements, with demographic parity difference decreasing by 88.31% and equalized odds difference dropping by 54.26%, both statistically significant ($p$-values < 1e-8). The work demonstrates the feasibility and importance of fairness-aware ML in aviation training datasets and points to extensions toward sensor-based monitoring and broader demographic attributes to enhance real-world applicability.
Abstract
While researchers have been trying to understand the stress and fatigue among pilots, especially pilot trainees, and to develop stress/fatigue models to automate the process of detecting stress/fatigue, they often do not consider biases such as sex in those models. However, in a critical profession like aviation, where the demographic distribution is disproportionately skewed to one sex, it is urgent to mitigate biases for fair and safe model predictions. In this work, we investigate the perceived stress/fatigue of 69 college students, including 40 pilot trainees with around 63% male. We construct models with decision trees first without bias mitigation and then with bias mitigation using a threshold optimizer with demographic parity and equalized odds constraints 30 times with random instances. Using bias mitigation, we achieve improvements of 88.31% (demographic parity difference) and 54.26% (equalized odds difference), which are also found to be statistically significant.