Research on the Application of SVM-RFE Algorithm in the Construction of Students' Academic Early Warning Model

This paper introduces the support vector machine-recursive feature elimination (SVM-RFE) algorithm to construct an academic early warning model for university students. This method makes use of the strong processing ability of SVM classification and the dimension reduction optimization of RFE. In data processing, the key factors that have a significant impact on students' academic performance are preliminarily classified and identified by SVM, then redundant features are eliminated by the RFE algorithm to reduce the complexity of the model. In this study, by collecting the real academic data of a college student, SVM-RFE algorithm is applied to feature screening and model construction, and its prediction performance is verified by experiments. The results show that the prediction accuracy of the model reaches 92.3%, and the false alarm rate is 7.8%. This further proves the effectiveness of the SVM-RFE algorithm in the application of academic early warning model.