Study Level Prediction of Indian and Hungarian Students towards ICT and Mobile Technology for the Real-Time

This paper focused on the identification of the level of student's study towards Information Communication Technology (ICT) and Mobile technology (MT) using Machine Learning (ML) algorithms. The level of study is classified into two groups such as Under Graduate (UG) and Post Graduate (PG) student. In the prediction of the above two groups, we used primary datasets with 4 supervised ML classifiers in IBM SPSS Modeler 18.2. The auto classifier algorithm suggested to modeled eXtreme Gradient Boosting Tree (XGBT), Random Tree (RT) and Random Forest (RF) out of other supervised ML algorithms. Also, Synthetic Minority Oversampling Technique (SMOTE) algorithm is applied to balance the dataset to enhance prediction accuracy. The findings of this paper conclude that the XGBT outperformed others in the study level prediction task. The XGBT classifier provided the highest accuracy of 92.36% with 16 significant features. On one hand, a significant difference was found between the accuracy of XGBT and RT and another hand accuracy makes no significant impact on RF and XGBT. The authors recommended these predictive models to be deployed as a realtime prediction of the study level of the university's student.