Predicting the academic progression in student's standpoint using machine learning

• Published in: Automatika Vol. 63; no. 4; pp. 605 - 617
• Main Authors: Sassirekha, M. S., Vijayalakshmi, S.
• Format: Journal Article Paper
• Language: English
• Published: Ljubljana Taylor & Francis 02.12.2022; Taylor & Francis Ltd; KoREMA - Hrvatsko društvo za komunikacije,računarstvo, elektroniku, mjerenja i automatiku; Taylor & Francis Group
• Subjects: Algorithms; Education; Feature selection; Machine learning; Prediction; student's performance; Students; Warning systems
• ISSN: 0005-1144; 1848-3380
• DOI: 10.1080/00051144.2022.2060652

Graduate students are unaware of their final qualification for a course. Even though there were many models available, few works with feature selection and prediction with no control over the number of features to be used. As a result of the lack of an improved performance forecasting system, students are only qualified on the second or third attempt. A warning system in place could help the students reduce their arrear count. All students undertaking higher education should obtain the qualification at their desired level of education without delay to transit to their careers on time. Therefore, there should be a predictive system for students to warn during the course work period and guide them to qualify in a first attempt itself. Although so many factors were present that affected the qualifying score, here proposed a feature selection technique that selects a minimal number of well-playing features. Also proposed a model Supervised Learning Approach to unfold Student's Academic Future Progression through Supervised Learning Approach for Student's Academic Future Progression (SLASAFP) algorithm that recommends the best fitting machine learning algorithm based on the features dynamically. It has proven with comparable predictive accuracy.