Research on Multi-level Student Achievement Analysis Method Based on Cluster Analysis

Abstract

Performance prediction can provide reference for teachers to improve teaching programs and students to improve learning methods. At present, most prediction methods use all students' grades to build prediction model, ignoring the multi-level characteristics of students. Therefore, a multi-level student achievement analysis method based on cluster analysis is proposed. Firstly, the sample data is clustered by affinity propagation clustering algorithm. Then, the prediction models are constructed for each sample. Finally, the corresponding prediction model is used to predict the performance. In order to verify the accuracy and efficiency of student achievement analysis, it is verified on the score data of college students of multiple majors. Through the experimental results we can see that the prediction accuracy of the Multi-level student achievement analysis algorithm based on cluster analysis is better.