Ensemble Feature Selection and Harris Hawks Optimization for Explainable Mental Health Risk Prediction in Female Sex Workers

High-accuracy prediction can identify at-risk individuals: The model achieved 95.78% accuracy in predicting depression among female sex workers, suggesting machine learning tools could help healthcare systems identify who needs mental health support before crises occur.

Violence and trauma are key risk factors: The analysis identified post-traumatic stress, client-related violence, and occupational factors as the major contributors to depression in this population, providing data-driven evidence for why targeted interventions addressing safety and trauma are critical.

Explainability matters for vulnerable populations: By using explainable AI methods, the researchers made the model's decision-making transparent—meaning healthcare providers can understand why the system flagged someone as at-risk, rather than treating it as a "black box," which is essential for building trust with marginalized communities.

Specialized models work better than generic ones: The hybrid approach outperformed traditional machine learning classifiers, suggesting that one-size-fits-all mental health prediction tools may miss important patterns specific to vulnerable groups like female sex workers who face unique stressors.

Technology can support early intervention: By combining accurate prediction with explainability, this tool could enable "early assistance, evidence-based targeted psychosocial care, and health planning" for a population that often faces barriers to accessing mental health services.