Predicting Student Performance in Online Learning During Pandemic Using Machine Learning Techniques

Any educational institution's main objective is to raise educational standards through the academic success of its students. To do this, schools can use Educational Data Mining (EDM), a growing field of research that builds on the core concepts of Data Mining (DM), to learn important facts like how students felt about the Covid-19 outage's impact on their online education. With EDM, educators have experimented with a variety of strategies for predicting students' behaviors in order to create optimal learning environments. Students' learning behaviors and academic outcomes can be predicted using techniques gleaned from the field of Educational Data Mining. The lockdown declared after the emergence of Covid-19 pushed several universities to embrace digital educational resources. Adoption of e-learning has expanded, as has the availability of data connected to online learning, in the decade after the COVID-19 epidemic. As a result, scientists have been motivated to create ML-based models to foretell students' achievements. This research study analyzes how the Covid-19 outbreak and its aftermath affected academic performance predictions. A standard dataset is used for this purpose; it is compiled online using a questionnaire and draws from a wide range of primary and secondary sources. In addition, four machine learning classification algorithms were employed to make predictions about the student's performance and then compared with state of art models. Researchers found that students' academic performance and mental health suffered when they used electronic learning aids like cellphones, computers, and tablets too much.