STRESS DETECTION USING MACHINE LEARNING

Abstract

Stress is a widespread health concern affecting both mental and physical well-being, with prolonged exposure linked to hypertension, cardiovascular diseases, sleep disturbances, depression, and neurological impairments. Traditional detection methods, such as self-report questionnaires and clinical interviews, are subjective, time-consuming, and unsuitable for continuous monitoring. Advances in machine learning (ML) and deep learning (DL) enable automated, real-time stress detection using physiological signals. Key indicators—including heart rate variability, respiration rate, snoring, blood oxygen levels, body movements, eye activity, and sleep patterns—reflect autonomic nervous system activity and reliably indicate stress. ML algorithms like Logistic Regression, Decision Trees, Random Forests, Support Vector Machines, Naïve Bayes, and K-Nearest Neighbors model relationships between physiological features and stress, while DL approaches such as Artificial Neural Networks, Convolutional Neural Networks, and LSTM networks learn complex patterns from sequential data. Combining these techniques allows accurate classification of stress levels, supporting continuous monitoring, early risk detection, and personalized interventions across healthcare, workplace management, and mental health domains.