In this research we have reviewed the performance of machine learning algorithms and WSNs for the monitoring student physical and mental health. We have found that logistic regression is a stronger alternative after an exhaustive evaluation of machine learning algorithms that consistently produce superior performance metrics than Gaussian mixture models (GMM). This has demonstrated average accuracy of 87.1% across 10 experiments, besting the average accuracy of 83.8% of GMM. It also illustrates greater efficacy for assessing system security, producing average accuracy of 86.6%, with 91.9% for GMM. The findings demonstrate the significance of select a machine learning algorithms that meets the specific needs of monitoring student health systems. They are also an illustration of logistic regressions dependability, and adeptness at identifying subtle changes in student health indicators that allow for proactive initiation of interventions and support mechanisms tailored to individual students’ needs. Going forward, monitoring systems should next be further refined and optimised so that they can more effectively support both student well-being and academic success in the complex and diverse educational settings of Indian universities.
Read full abstract