Using Machine Learning Algorithms to Predict Students’ Performance and Improve Learning Outcome: A Literature Based Review
Loading...
Date
2020-03
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
Keywords
Machine learning algorithms, students’ performance, learning outcome
