A Machine Learning Approach for Predicting the Anxiety Status Among O/L Students: A Case Study from Elpitiya Educational Zone
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
In Sri Lanka, anxiety and depression are common among youth. This study examined anxiety among$\mathrm{O} / \mathrm{L}$students in the Elpitiya Educational Zone, aiming to identify influential factors and apply machine learning for early detection. After getting the IOBSL ethical clearance approval using the SCARED child questionnaire and a stratified cluster sampling method, data were collected from 417 students, representing a population of 7,060. The study constructed one main response variable (Anxiety), and five sub-anxiety categories are Panic Disorder, Generalized Anxiety, Social Anxiety, Separation Anxiety, and School Avoidance. Overall,$\mathbf{6 1. 3 \%}$of respondents experienced some level of anxiety, 42.2 % with possible anxiety, and 19.1 % with specific anxiety. The Random Forest classifier yielded the best performance for predicting anxiety, with an accuracy of 63.33 %, an$F 1$-score of 0.6239, a recall of 0.6333, and a precision of 0.6710. For multi-label sub-anxiety classification, a Multioutput Classifier with Random Forest achieved an F1-score of 0.7186 and a Hamming loss of 0.3017. Significant factors influencing anxiety included gender, BMI, screen time, physical activity, sleep duration, parental employment, school type, and academic stress. Notably, female students, those under high academic pressure, and those from smaller schools are more prone to anxiety, while regular exercise and 5-7 hours of sleep help reduce it. The results support the importance of mental well-being through awareness and early intervention.
Fields of Study & Subjects
Artificial Intelligence in Healthcare
Student Stress and Coping
Child and Adolescent Psychosocial and Emotional Development
Publication details
- Journal / source
- International Conference on Advanced Research in Computing (ICARC)
- Pages
- 1–6
- Publication date
- 18 February 2026
- Language
- EN