基于BP神经网络与决策树的大肠癌虚实证型分类对比

Comparison on Classification of Excess and Deficiency Syndromes of Colorectal Cancer Based on BP Neural Network and Decision Tree

采用二进制编码对症状数据进行量化,将专家归纳的8个证型分为虚实两证并赋值量化,建立基于BP神经网络与决策树的大肠癌虚实证型分类模型,结果显示BP神经网络分类模型较决策树分类模型更适合于非线性映射关系的处理。

The paper quantizes symptom data through binary coding,divides 8 syndromes summed up by experts into excess and deficiency syndromes,values and quantizes them,and establishes the model for classification of excess and deficiency syndromes of colorectal cancer based on BP neural network and decision tree.The result shows that BP neural network classification model is more applicable for the handling of the nonlinear mapping relation compared with decision tree classification model.