Predicting student learning outcomes based on supervised machine learning model

Abstract

Predicting student learning outcomes is an important area of research in the field of education. The purpose of this paper is to propose a learning machine model that can reliably predict students' learning outcomes. The model is based on a set of attributes related to students such as family, school, and previous learning outcomes. The experimental results of the paper show that the Random Forest algorithm is very effective in predicting the learning outcomes of with a Mean Absolute Error (MAE) of 1.13. Besides, attributes such as family size, student age, school and reasons for choosing a school are important factors affecting the prediction results. The proposed predictive model could help educators identify students at risk of low academic achievement and take appropriate remedial measures to improve their academic performance. Furthermore, predicting academic performance also helps students and
parents make the necessary adjustments to improve academic achievement.