Measuring Psychometric Analysis of Stress Level in Learners Across Multiple Parameters Using Machine Learning

Abstract

This study presents a comprehensive exploration into the analysis of stress levels in learners using machine learning methodologies across diverse parameters. The study investigates stress quantification through multi-dimensional data encompassing different factors. Employing machine learning models, the study achieves an overall predictive accuracy of 85.6% in assessing stress levels. Notably, physiological data analysis yielded an accuracy of 88.9%, highlighting the reliability of identifying stress patterns. Furthermore, a strong negative correlation -0.75 between stress levels and performance was observed, indicating a significant impact of stress on learners' educational outcomes. Environmental factors contribute to 28% of the variability in stress levels in learners, underscoring their influence. Noteworthy features in predicting stress levels include heart rate variability 37.5%, sleep quality 23.8%, and social interactions 18.6%.