Machine Learning Predictive Model of Academic Achievement Efficiency based on Data Envelopment Analysis

Abstract

Along the way with the changes in the education landscape nowadays, the grade is not the only determinant to predict the students' success. In the context of a student's academic performance, it is better to focus on measuring the efficiency of academic achievements that used multiple determinants of holistic outcome rather than just focus on the student grade. Data Analysis Envelopment (DEA) is a nonparametric method that widely used in many fields to measure performances efficiency but limited research has been reported on DEA in education domain. Acknowledging DEA time consuming issue when involving a huge size of data, recent research on deploying machine learning in DEA keeps on rapid progressing. This paper presents a new research framework of DEA and Auto-ML predictive model for the academic achievement efficiency. The framework includes variety options of machine learning to be compared from the conventional manual setting into the recent Auto-ML technique. The research framework will provide new insights into the decision-making process particularly in the education context.