REAL-TIME ANALYSIS OF STUDENT BEHAVIOR FOR IDENTIFYING SOCIAL ANXIETY IN HIGH SCHOOL USING MACHINE LEARNING

Abstract

In this study, high school students at Little Scholars Matriculation Hr. Sec. School in Thanjavur, Tamil Nadu, India, are asked about the prevalence and effects of social anxiety. The 11-item Social Phobia Inventory (SPIN) questionnaire, which asks about social interactions, fear of being judged, and discomfort in different social settings, was used to collect data from students. The study analyses student replies and determines the degree of social anxiety using this dataset and a Random Forest machine learning technique. Through the identification of important characteristics that lead to higher degrees of discomfort, the model seeks to predict social anxiety levels. By use of feature selection and correlation analysis, the research reveals intricate connections among many facets of social interactions that impact anxiety. Accuracy and predictive power are used to assess the Random Forest model's performance, showing that it can accurately predict high school students' social anxiety. In order to improve mental health support systems for high school kids, the study suggests more research to improve predictive models and emphasises the potential of Random Forest for precisely pinpointing important elements linked to social anxiety