Students’ Classification Model Based on Stacking Algorithm

Abstract

A model for Students’ Classification which aims to help college teachers teach in accordance with students’ aptitude is proposed based on the daily practice data of a course. Based on the Stacking method, this classifier model ensembles eXtreme Gradient Boosting algorithm, Random Forest algorithm and the Logistic Regression algorithm. The experimental results show that this model can divide effectively students into different classes and the accuracy can be significantly improved by comparing with other classifier models. And it can serve as reference and guidance for college students’ classifier on other courses.