Predicting the Performance Fluctuation of Students Based on the Long-Term and Short-Term Data

Abstract

The potential value of students' academic performance prediction has been extensively studied by educational institutions. However, it still has great research challenges, such as the relationship between students' behavior and their academic performance. This paper reused data from online educational platforms, and used four methods to analyze educational value in relation to the fluctuations of academic performance. The methods are based on Step Regression, Logistic Regression, Decision Tree and Support Vector Machine Regression (SVMR). At last, SVMR model is selected by comparing the prediction accuracy of the four models. The experimental results show that there are some differences between traditional cognitive performance and prediction and that educational decisions can be driven by the data.