Analysis of the Impact of Artificial Neural Networks on Predicting Students' Academic Success

Abstract

The use of Artificial Neural Networks (ANNs) has become increasingly prevalent in educational research, particularly in predicting students' academic success. This article provides a comprehensive review and descriptive analysis of studies conducted on the application of ANNs in forecasting educational outcomes. The review covers the historical evolution of ANNs, their applications in education, and specifically their use in predicting academic success. Through the analysis of various studies, it was found that ANNs outperform traditional predictive methods due to their ability to process complex, multidimensional data. The results indicate that ANNs can accurately predict student performance by analyzing a wide range of variables, including participation in class activities, past academic records, and demographic factors. Despite their potential, the implementation of ANNs faces challenges such as the need for large, high-quality datasets and the complexity of interpreting the results. This study highlights the current trends, common patterns, and limitations in using ANNs for educational purposes. It concludes with recommendations for future research, emphasizing the need for further exploration of deep learning models and improved interpretability of ANN outcomes. The findings of this study can be practically applied in educational settings to enhance student performance prediction, support decision-making in educational policy, and ultimately improve the quality of education.