Predictive Model of Undergraduate Student Grading Using Machine Learning for Learning Analytics

ABSTRACT

Predicting the academic achievement has become very important for university students as well as lecturers, especially during the difficult times of pandemic Covid-19, online distance learning (ODL) with some students need to do part-time job due to financial problems. Furthermore, we are surrounded by a plethora of digital entertainment, such as social media platforms and mobile games, which may also serve as a distraction and undermine students' commitment to their studies. Therefore, this paper develops a model to predict student academic performance using Machine Learning approaches. The model is developed by training the dataset acquired that consists of demographic information, study preparation, academic performance and motivation from the students in a public higher institution in Malaysia. The model is also tested on the public dataset related to student academic performance. The findings showed that JRip classifier have obtained the best accuracy of 92% for the newly collected data and 100% accuracy by using Random Forest classifier on the public dataset. The developed model and data visualization are useful for the development of learning analytics system which students and lecturers can make an early intervention and determining whether students need to take necessary actions to improve their academic results in real-time, as well as gaining a better understanding of the factors that may affect their academic performance. © 2022 IEEE.