Using Machine Learning Algorithms to Predict Students’ Performance and Improve Learning Outcome: A Literature Based Review
Loading...
Date
2020
Authors
Ofori, Francis
Maina, Elizaphan
Gitonga, Rhoda
Journal Title
Journal ISSN
Volume Title
Publisher
Stratford Peer Reviewed Journals and Book Publishing
Abstract
The application of machine learning techniques in predicting students’ performance, based on their background and their in-term performance has proved to be a helpful tool for foreseeing poor and good performances in various levels of education. Early prediction of students’ performance is useful in taking early action of improving learning outcome. The prediction of the student's academic performance is important as it helps increase graduation rates by appropriately guiding students, guiding changes in university academic policies, informing instructional practices, examining efficiency and effectiveness of learning, providing meaningful feedback for teachers and learners and modifying learning environments. A high prediction accuracy of the students’ performance is helpful to identify the low performance students at the beginning of the learning process. However, to achieve these objectives, large volume of student data must be analyzed and predicted using various machine learning models. Moreover, it is not clear which model is best in predicting performance and which machine learning model is appropriate in improving learning in among students. The paper through intensive literature review attempts to identify best machine learning model in predicting student performance and appropriate machine learning model in improving learning. The empirical review indicated contentious results on machine learning model that best predicts students’ performance. Moreover, it is not clear among the various machine learning algorithms which one derives the best approach in predicting students’ performance while improving learning outcome. The varying prediction level by various machine learning models may be as a result of differences in socioeconomic. It may also be important to note that student’s academic performances are affected by many factors, like socioeconomic factors of students like family income, parental level of education and employment status of students or parents but are not considered when testing the accuracy of various machine learning models in predicting students’ performance. Moreover, the various machine learning models did not identify the most appropriate machine learning model in improving students’ outcome. Most models focused largely in predicting students’ performance without considering mechanisms to improve learning outcome of students. As a result, it is important to test the accuracy of various machine learning models that best predicts students’ performance and the one that is most appropriate in improve learning outcome while considering socio economic and demographic factors of the students. The study makes a conclusion that predicting students’ performance is of the highest priority for any learning institution across the globe. Using various machine learning methods to accurately predict student’s performance would be highly required. It is important to accurately rank machine models based on their prediction capabilities in predicting students’ performance and in improving learning outcome.
Description
A research article published in Journal of Information and Technology
Keywords
Machine learning algorithms, students’ performance, learning outcome
Citation
Ofori, F., Maina, E. & Gitonga, R. (2020). Using Machine Learning Algorithms to Predict Students’ Performance and Improve Learning Outcome: A Literature Based Review, Journal of Information and Technology, Vol. 4(1), 33-55.