Simulation of Machine Learning Techniques to Predict Academic Performance

Abstract

Educational Data Mining (EDM) is that branch of Artificial Intelligence (AI) which uses a combination of Data Mining (DM) and Machine Learning (ML) techniques to make predictions on aspects specifically related to students, teachers, and, to educational institutions, in general. The goal of any educational institution is to ensure that every student gets the “right” foundational, high-quality education according to its inherent and acquired talents and potential; in order to provide the correct direction for the student's future education and successful career. Predicting a student's academic performance in advance, is therefore imperative to enable educational institutions and parents to take proactive decisions to steer students in the correct direction through Student Intervention Strategies. This paper simulates the most commonly used Supervised Machine Learning Algorithms that have been used for predicting Academic Performance namely, Decision Tree (DT), Random Forest (RT), K-Nearest Neighbors (KNN), Logistic Regression (LR), Naïve Bayes (NB), Gradient Boost Tree (GBT), Multi-Linear Perceptron (MLP) and Support Vector Machine (SVM). The objective of this paper is to analyze these classification algorithms for EDM in depth, in order to determine the most suitable attributes and the most efficient Machine Learning Model to accurately predict the academic performance of students and ensure academic success.