Application of Advanced Statistical Techniques to Improve the Prediction of Student Performance in Mathematics

Abstract

The present study explores the application of advanced statistical techniques to predict the academic performance of students in the area of mathematics. Through the use of logistic regression models, decision trees, and neural networks, data from 500 high school students in public institutions were analyzed. The results show that advanced statistical techniques allow a more accurate prediction of academic performance, with a success rate of 87% in neural network models. These findings suggest that the integration of these tools can facilitate the early identification of students at risk of low achievement and improve educational interventions.