Simulation-Based Machine Learning for Predicting Academic Performance Using Big Data

Abstract

In this study, simulation and big data analytics are combined with machine learning techniques, specifically K-means clustering, Apriori algorithm, and a stacked integrated learning model, to predict academic performance of college students with a high accuracy of 95.5%. By analyzing behavioral data from over 1,000 undergraduates, we correlate various behaviors with academic success, focusing on the use of libraries, self-study habits, and internet usage. Our findings highlight the benefits of using big data and simulation in educational strategies, promoting effective resource allocation and teaching enhancements. The study acknowledges limitations due to its regional focus and proposes future research directions to enhance model generalization and technological integration for broader application.