Stress Detection through EEG Signals: Employing a Hybrid Approach integrating Time Domain, Frequency Domain Features and Machine Learning Techniques

Abstract

Mental stress is a widespread problem that affects people all over the world and can have negative health consequences if not appropriately controlled. The importance of early detection and management in minimizing the detrimental effects of stress cannot be overstated. This study provides a hybrid technique for stress detection that combines time domain and frequency domain information taken from EEG data. Machine learning techniques are used to create a stress detection model that is accurate and dependable. The goal is to increase the accuracy and reliability of stress detection so that prompt intervention and assistance may be provided. The paper opens with an introduction to stress, its effects on mental health, and the need for automated stress detection systems. EEG signals are introduced as a valuable data source for recording stress-related brain activity. To test the model's success in stress detection, performance evaluation criteria such as accuracy, sensitivity, specificity, and F1-score are used. When the result compared to the present approach, the Hybrid Approach has higher accuracy (SVM-98.33% & RF-95%).