Learning analytics using deep learning techniques for efficiently managing educational institutes

Abstract

Increasing numbers of higher education institutions see themselves as service providers, catering primarily to the needs of its students. The improvement of student performance is a top priority for universities. It is critical to first assess the present situation of the students before designing a program to improve their performance. Higher education administrators face a significant problem in predicting a student's future success. The goal of this study is to learn what factors influence college students' decision on a major. It will be possible to forecast students' behavior, attitudes, and performance with the use of predictive tools and procedures. Predicting student performance ahead of time makes it possible to take proactive measures to raise achievement levels. To obtain a high education standard, several attempts have been made to forecast student performance. However the accuracy of these predictions falls short of the desired level of excellence. Machine learning approaches including Artificial Neural Network, Nave Bayes, and SVM are being studied. A University Data Set from UCI Machinery is used in the experimental investigation.