INTERNATIONAL JOURNAL OF ADVANCE RESEARCH
IN MULTIDISCIPLINARY

Stress poses major risks to mental and physical health. Electrocardiogram (ECG) signals, which reflect the heart's electrical activity, offer a non-invasive approach to stress detection. This study proposes a machine learning-based method for classifying stress levels using ECG signals, leveraging support vector machines (SVM) and neural networks. By automating feature extraction and classification, the model achieves improved accuracy and adaptability over traditional handcrafted approaches. The system is evaluated through a comprehensive pipeline including data collection, preprocessing, feature selection, and classification, demonstrating high performance and clinical potential.