目的:运用机器学习法构建社会支持对大学生心理健康影响的预测模型。方法:采用方便抽样法,选取2024年1月至2024年6月在武汉市高校就读的在校大学生开展调查。调查工具包括一般情况调查表以及社会支持量表和症状自评量表。使用多因素线...目的:运用机器学习法构建社会支持对大学生心理健康影响的预测模型。方法:采用方便抽样法,选取2024年1月至2024年6月在武汉市高校就读的在校大学生开展调查。调查工具包括一般情况调查表以及社会支持量表和症状自评量表。使用多因素线性回归分析社会支持是否为大学生心理健康的影响因素,构建随机森林模型计算模型的稳健性以及影响因素的权重。结果:共回收问卷520份,有效问卷503份,问卷有效率为96.7%。回归模型结果表明,社会支持(t = −3.656, p = 0.000 Objective: To construct a predictive model of the impact of social support on college students’ mental health by using machine learning. Methods: A convenient sampling method was used to select college students in Wuhan from January 2024 to June 2024. The survey tools included the general situation questionnaire, the social support scale and the symptom self-rating scale. The multifactor linear regression model was used to analyze whether social support was an influencing factor for college students’ mental health, and the robustness of the model and the weight of the influencing factors were calculated by building a random forest model. Results: A total of 520 questionnaires were collected, 503 were valid, and the effective rate was 96.7%. The results of regression model showed that social support (t = −3.656, p = 0.000 < 0.01) and other six factors had statistical significance on the mental health level of college students. The results of random forest model showed that the model predicted better, with a Mean Square Error (MSE) of 0.365. The top three weight shares were social support, monthly family income, and whether or not smoking. Conclusion: Social support is an important factor affecting college students’ mental health. It is suggested that parents pay attention to their children’s school life, and the school promotes student exchanges. Regular psychological counseling activities should be carried out to cultivate a healthy lifestyle, reduce the academic burden and reduce the impact of stress.展开更多
文摘目的:运用机器学习法构建社会支持对大学生心理健康影响的预测模型。方法:采用方便抽样法,选取2024年1月至2024年6月在武汉市高校就读的在校大学生开展调查。调查工具包括一般情况调查表以及社会支持量表和症状自评量表。使用多因素线性回归分析社会支持是否为大学生心理健康的影响因素,构建随机森林模型计算模型的稳健性以及影响因素的权重。结果:共回收问卷520份,有效问卷503份,问卷有效率为96.7%。回归模型结果表明,社会支持(t = −3.656, p = 0.000 Objective: To construct a predictive model of the impact of social support on college students’ mental health by using machine learning. Methods: A convenient sampling method was used to select college students in Wuhan from January 2024 to June 2024. The survey tools included the general situation questionnaire, the social support scale and the symptom self-rating scale. The multifactor linear regression model was used to analyze whether social support was an influencing factor for college students’ mental health, and the robustness of the model and the weight of the influencing factors were calculated by building a random forest model. Results: A total of 520 questionnaires were collected, 503 were valid, and the effective rate was 96.7%. The results of regression model showed that social support (t = −3.656, p = 0.000 < 0.01) and other six factors had statistical significance on the mental health level of college students. The results of random forest model showed that the model predicted better, with a Mean Square Error (MSE) of 0.365. The top three weight shares were social support, monthly family income, and whether or not smoking. Conclusion: Social support is an important factor affecting college students’ mental health. It is suggested that parents pay attention to their children’s school life, and the school promotes student exchanges. Regular psychological counseling activities should be carried out to cultivate a healthy lifestyle, reduce the academic burden and reduce the impact of stress.
