Prediction of the Effects of Online Education on Students Health by Machine Learning

Abstract

The Covid-19 pandemic has disrupted educational activities across Bangladesh, resulting in the closure of educational institutes, affecting nearly four crore students' daily learning activities. Besides, the economic downturn brought on by the pandemic is impacting students and their families. Access to educational opportunities is increasingly difficult for students from low-income families, and the rising poverty rate adds to their woes. According to various projections, Bangladesh's poverty rate is estimated to be around 35-40%, compared to a pre-pandemic rate of around 20%. Poverty has a disproportionate impact on lower-income families, where survival takes precedence over schooling. Otherwise, most of the students are having trouble with understanding the new online platform. Not only that, the poor students cannot afford a laptop to attend classes. Internet issues are also a massive problem, as most of the students have moved from cities to villages to their hometowns, they cannot attend the classes because of low internet connectivity. Those who are attending classes online and came familiar with the online platform, most of them have become addicted to the internet. As a result, they are hampering their eyesight and other physical problems. We have applied seven traditional ML algorithms. They are- KNN, Naïve Bayes, Decision Tree, Support Vector Machine (SVM), Neural Network algorithm, Random Forest, and AdaBoost. The best accuracy was predicted by the neural network, which was 85.45%