论文部分内容阅读
目的以麻疹暴发突发公共卫生事件为例,探索研究突发公共卫生事件分级量化判定指标、量化判定模型及对模型应用的评价。方法使用SPSS Clementine 11.1.1软件,采用C5.0决策树与多分类有序反应变量的logistic回归分析建立模型。结果通过C5.0算法获得了决策树模型以及自变量指标的重新分组的标准,并进一步进行多分类有序反应变量的logistic回归分析,模型评价结果显示两者对于麻疹突发事件的级别判定和预测都具有很高的正确性。结论结合应用决策树模型和logistic回归模型可实现对麻疹突发事件自动进行级别判定的目的。
Objective To take the outbreak of public health emergency measles as an example, to explore the classification of public health emergency quantitative evaluation indicators, quantitative decision model and evaluation of the model application. Methods SPSS Clementine 11.1.1 software was used to establish a model using logistic regression analysis of C5.0 decision trees and multi-class ordered response variables. Results The decision tree model and the regrouping criteria of independent variables were obtained by C5.0 algorithm. The logistic regression analysis of multi-classification and orderly response variables was further carried out. The results of model evaluation showed that both the level of determination of measles emergency and Forecasts are highly accurate. Conclusion The purpose of judging the incident level of measles automatically can be achieved by combining the application of decision tree model and logistic regression model.