Impact of Artificial Intelligence and Machine Learning on Predicting Student Performance and Engagement

Abstract

The emergence of Artificial Intelligence (AI) and Machine Learning (ML) has improved the way education systems can predict and in turn effectively improve the student performance levels and activity level. This paper aims at analyzing the effectiveness of integrating the AI/ML models in handling educational data to predict academic performance, and monitor student interactions. Using Decision trees, Neural trees and ensemble methods, various important parameters like grades, attendance, participation and behavior indicators are predicted accurately. In the study, a dataset was obtained comprising records of 15000 students across various educational institutions to train and validate the models. The analysis identified that average prediction accuracy of AI/ML algorithms was 92% for academic results and 88% for engagement indicators, which are better than statistical methods. Predictors such as studying, time management and participation in co-curricular activities were among some of the features that needed consideration, according to feature importance analysis. Furthermore, early warning system models developed by ML models enable timely intervention hence, meaning that dropout levels were cut by 25%. This research further explores the alternate reality where AI/ML can revolutionise education systems by offering insights to educators, effective pathways of learning as well as prevention strategies. However, issues like data security, the model’s inherent bias, and the level at which it can be scaled present some limitations for wider adoption.