基于Logistic回归与决策树的心脏病确诊因素分析

Analysis of Diagnostic Factors of Heart Disease Based on Logistic Regression and Decision Tree

通过利用加州大学欧文分校数据库的心脏病数据集,通过建立logistic模型和决策树模型分析确诊心脏病的危险因素。采用ROC曲线和AUC面积作为标准来评价模型预测效果,结果显示两种模型对于数据的拟合都表现不错。与此同时两种模型显示胸痛类型、静息血压、荧光染色法测定的主要血管数和是否患地中海贫血症对于最终是否确诊心脏病有显著影响。

By using the heart disease data set of the University of California Irvine database,the risk factors of diagnostic heart disease are analyzed by establishing the Logistic Regression model and Decision Tree model.The ROC curve and AUC area are used as criteria to evaluate the prediction effect of the model.The results show that the two models perform well in fitting the data.At the same time,the two models show that the type of chest pain,resting blood pressure,the number of main blood vessels measured by fluorescent staining and whether or not suffering from thalassemia has a significant impact on whether the final diagnosis of heart disease occurs.