决策树模型在Ⅱ型糖尿病诊断中的应用

Application of Decision Tree Model on Type Ⅱ Diabetes

根据临床检验资料信息,利用决策树模型建立Ⅱ型糖尿病诊断模型.首先搜集数据并进行预处理,采用C4.5和CART算法构造分类模型,然后比较两种方法的准确率,最后利用受试者工作特征曲线(ROC)的评价方法对预测结果进行评估.为更准确地诊断Ⅱ型糖尿病提出理论依据.

According to clinical test data, the diagnosis of type Ⅱ diabetes is established using decision tree model. Firstly, the data is collected and preprocessed; the classification model is constructed by using C4. 5 and CART algorithm. Then the accuracy of both methods is compared. At last, the prediction result is assessed by using ROC evaluation method. For more accurate diagnosis of type Ⅱ diabetes, a theoretical basis is put forward.