This is the authors' abstract. We don't add key points for this paper.
Adolescent stress has emerged as an important concern due to its impact on mental well-being, academic performance, and long-term health issue. Wearable sensing devices and machine learning pattern provide opportunities for objective and continuous stress monitoring techniques. This study aims to develop a machine learning–based framework for detecting stress levels in adolescents using wearable devices. We utilized the Fitbit Sense 2 smartwatch to record several physiological metrics from adolescents aged 12 to 16, specifically targeting their heart rate, electrodermal activity (EDA), heart rate variability (HRV), and physical movement levels. After preprocessing, feature engineering and z-score–based composite stress scores were used to generate low, moderate, and high stress labels.Experimental results showed that RF and Gradient Boosting achieved the highest average accuracy of 94.78%, with a peak accuracy of 97.61%. Electrodermal activity and heart rate–based features emerged as significant predictors of stress. The study confirms that wearable-derived biological signals are highly effective for stress identification, indicating a significant opportunity to use artificial intelligence for teenage mental health surveillance. Future work may extend this framework through multimodal integration of psychological measures and physiological wearable data for more comprehensive stress assessment.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Rane et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: