Supervised ML based Model to Predict Satisfaction of Student Performance During the Pandemic

The coronavirus, COVID-19 has spread around the globe, causing unprecedented changes in major areas including education and the economy. Amid a pandemic, the educational sector is relying heavily on performance prediction models to aid in the study of student status and the subsequent improvement of performance. The term “Educational Data Mining” describes the practice of using data mining principles and methods in the classroom. The explosion of educational data in recent years has made educational data mining a hot issue. Methods for discovering latent patterns in educational data might be developed within this area. The data can be learned more about student’s, their learning habits, and how well they will do in school by applying the techniques drawn from Educational Data Mining. This study aims to predict how universities would modify e-learning in the aftermath of the outbreak by investigating student’s actual experiences utilizing e-learning facilities during this time. A formal questionnaire was distributed and an online survey was created for this purpose. Following an exhaustive survey, a good outcome was achieved in terms of the dataset. In addition, the selected supervised machine learning algorithms such as Multi-Layer Perceptron (MLP), Random Forest (RF), Support Vector Machine (SVM), and XG Boost (XGB) are utilized to train the prediction model to determine how likely students are to adopt the e-learning approach in the future. MLP produced the best results compared to other approaches, with an accuracy of 84.1%. The unexpected emergence of the epidemic caused tremendous disruptions to the academic pursuits of people.