An Unsupervised Machine Learning Model for Discovering the Impact of Social Networks on Students’ Behavior and Health in a COVID-19 Case Study

This chapter investigates the impact of social networks on students’ behavior and health during COVID-19 using unsupervised machine learning techniques. The data were collected from an online survey of 1182 students and preprocessed for analysis. The SimpleKMeans clustering algorithm was used to identify six distinct clusters of students based on age, experience with remote online learning, preferred social networking sites, weight changes during the pandemic, health issues, time spent on social networks etc., and connection to others. The clustering results revealed significant differences in students’ behavior and health based on their social network usage patterns. Specifically, students who spent more time on social networks and had more connections reported higher stress levels, anxiety, and decreased physical activity during the pandemic. The findings have important implications for education and public health policies aimed at promoting healthy behavior and reducing the negative impact of social media on students’ well-being. Future research can explore additional factors influencing social network usage and its impact on health outcomes.