Predicting Student Academic Path Using Machine Learning: Systematic Review

Machine learning have brought a great revolution for a multitude of industries. In fields like education, ML algorithms offer potential for transforming traditional approaches by improving the monitoring and supervising student progress and identifying learning orientations. While shaping the educational experience of students, predicting academic path has also an essential role for educational institutions. This knowledge enables effective career guidance and allows students to make informed decisions about their education and career paths early on. This article delves into the prediction studies proposed in between 2011 and 2023. The first step was conducting deep researches to identify relevant publications. This search identified a total of 91 publications. Subsequently, each publication went through meticulous evaluation against specific criteria established for the synthesis process. Following this rigorous screening process, 42 publications were found and selected. Across the spectrum of studies analyzed, researchers showcased a diverse range of ML techniques used to forecast student’s performance and anticipate their future academic trajectories. These techniques span from traditional algorithms such as decision trees and logistic regression to more advanced models including neural networks and ensemble methods. Furthermore, many studies concentrated on feature selection methodologies to augment the accuracy and robustness of their analyses. The conclusions reflect that Machine Learning methods have been shown to be crucial in predicting students pathway, thus improving the students’ performance.