基于非线性SVM的高职学生心理预警模型研究

打开文本图片集

中图分类号：TP391 文献标识码：A 文章编号：2096-4706（2025）09-0162-06

Abstract： This research addresses the psychological health issues of vocational college students， collects multi-dimensional data on students' psychological states， personal information， and life behaviors， and uses the nonlinear SVM algorithm to construct a psychological early warning model. The model can effectively classify and identify student psychological crises into four levels of severe psychological crises， general psychological issues， potential psychological distress， and no psychological distress. The empirical research design includes data collection and preprocessing， feature weight normalization processing， and data analysis. The results show that characteristics such as psychological state， family composition， and family economic status have a significant impact on student psychological crises. The SVM model achieves an early warning accuracy of up to 93.5% ， which is superior to models such as Multilayer Perceptron and Random Forest， providing a practical tool for mental health screening and intervention among vocational college students.

Keywords： early warning model; colleges and universities; psychological health; SVM algorithm

0 引 言

目前，在我国的高等职业学院中，学生的心理健康现已成为全社会共同关注的焦点，其状态对学生的成长健康与校园的整体稳定与安全至关重要[1]。（剩余6671字）