Aggregating Time Series and Tabular Data in Deep Learning Model for University Students’ GPA Prediction

Abstract

Current approaches of university students' Grade Point Average (GPA) prediction rely on the use of tabular data as input. Intuitively, adding historical GPA data can help to improve the performance of a GPA prediction model. In this study, we present a dual-input deep learning model that is able to simultaneously process time-series and tabular data for predicting student GPA. Our proposed model achieved the best performance among all tested models with 0.4142 MSE (Mean Squared Error) and 0.418 MAE (Mean Absolute Error) for GPA with a 4.0 scale. It also has the best R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> -score of 0.4879, which means it explains the true distribution of students' GPA better than other models.