Optimized Predictive Models in Health Care Using Machine Learning

20StressDetect: ML for Mental Stress Prediction

Himanshu Verma¹, Nimish Kumar²*, Yogesh Kumar Sharma³ and Pankaj Vyas¹

¹Manipal University, Jaipur (Rajasthan), India

²BK Birla Institute of Engineering and Technology, Pilani (Rajasthan), India

³KLE Foundation, Greenfield, Vaddeswaram, Guntur (Andhra Pradesh), India

Abstract

Mental stress has become a growing concern in recent years, negatively impacting physical and psychological health. This chapter presents a methodology that relies on machine learning to forecast and categorize mental stress levels based on physiological signals. The dataset used include, including an electrocardiogram (ECG), electrodermal activity (EDA), and respiration signals obtained from participants under stressful conditions.The sign was preprocessed to eliminate noise and artefacts, and feature extraction techniques were applied to obtain relevant features for stress prediction. Three distinct machine learning algorithms were employed to classify stress levels: Random Forest (RF), Support Vector Machine (SVM), and Artificial Neural Network (ANN). Evaluation metrics were used to measure the performance of each algorithm, and an experimental setup was designed to compare their performance.

The results show that all three algorithms accurately predicted and classified stress levels. Out of the three algorithms evaluated, the Random Forest algorithm demonstrated the best performance in terms of accuracy, with the Support Vector Machine and Artificial ...