logistic回归分析和BP神经网络模型在心理健康状况预测中的应用

Comparison between logistic regression analysis and BP neural network model in aspects of psychological stress

目的通过logistic回归分析与BP神经网络模型在心理应激相关影响因素研究的结果对比,建立心理应激的神经网络模型。方法对某医学院校785名本科新生进行心理应激以及相关影响因素问卷调查。结果BP神经网络模型对心理健康类的正确预测率为99.7%,对有心理问题类的正确预测率为95.1%,总的符合率为98.5%。logistic回归分析对心理健康类的正确预测率为94.1%,对有心理问题类的正确预测率为75.0%。总的符合率为89.2%。结论人工神经网络模型比logistic回归分析有更好的预测功能。

Objective To establish a neural network model of psychological stress by comparing the logistic regression analysis and BP neural network model on the influencing factors of psychological stress results. Methods QA questionnaire survey was conducted among 785 freshmen for their psychological stress and related problems. Results The correctly predict rate of the BP neural network model to the mental health was 99.7%, the rate to the mental problems was 95.1%, and the total coincidence rate was 98.5%. The correctly predict rate of the logistic regression analysis to the mental health was 94.1%, the rate to the mental problems was 75.0%, and the total coincidence rate was 89.2%. Conclusion The artificial neural network model has a better prediction function than the logistic regression analysis.