RETRACTED ARTICLE: Towards developing an ensemble based two-level student classification model (ESCM) using advanced learning patterns and analytics

Abstract

In recent decade, learning analytics has gained more attention and several advanced data mining models are developed for deriving the hidden sources from educational databases. The extracted data helps the Educational Institutions or Universities to enhance the teaching methodologies of faculties and student’s learning process in efficient manner. For improving the student performance and better educational results, the student data evaluations based on their learning modes are significant. With that note, the proposed model develops a new model called ensemble based two-level student classification model (ESCM) for effectively analysing and classifying the student data. With the student data pursuing technical higher education, the ESCM is performed with three traditional classification model and ensemble classifier techniques for enhancing the classification accuracy. The model utilizes support vector machine, Naive Bayesian and J48 classifier that are combined with Ensemble classification methods as modified meta classifier such as bagging and Stacking. Here, the technical higher education student data collected from SRM student database based on the feature set contains the student learning factors that support performance enhancement. The results are evaluated with the SRM student datasets and compared based on the classification accuracy and model reliability. Furthermore, the obtained results outperform the existing models. Based on the accurate predictions, special attentions and measures are taken to improve the student results and institutional reputation.