An explainable artificial intelligence-based approach for prediction and analysis of mental health disorder due to work pressure in the tech industry

作者：Vaishali Mehta Wadhwa, T. Jena, Jyoti Garg · 发表于：Journal of Information and Optimization Sciences · 年份：2025 · DOI：10.47974/jios-2011 · 被引用次数：2

These days, mental health disorders are becoming a major problem, particularly for workers. Effective management and treatment of mental health illnesses depend on early and precise diagnosis. This study investigates how to enhance the interpretability of machine learning models in the prediction of mental illness by utilizing explainable artificial intelligence (XAI) techniques, namely SHAP (Shapley Additive Explanations) and LIME (Local Interpretable Model-agnostic Explanations). We used a dataset from a survey of IT companies that had the required variables in order to create a prediction model. Our results showed that important variables like age, privacy, and work-related distractions had a big impact on model predictions. SHAP values shed emphasis on the role of individual characteristics, emphasizing the significance of age, work interference, and privacy as key risk factors for mental illness. The dependent graphs showed a strong relationship between key characteristics and model output, highlighting the necessity of employee diagnosis and preventative care. The findings highlight the potential of XAI techniques in therapeutic settings, opening the door to more individualized diagnosis and better outcomes for patients with mental health conditions.