属性层次模型在中医证候诊断中的应用

目的：研究一种综合考虑专家主观经验和机器学习客观规则的中医证候诊断模型。方法：基于属性层次分析模型，提出了一种可以应用于中医证候识别的方法，并编程实现了智能系统。系统融合中、西医多个特征构建决策属性，集成中医专家主观经验和机器学习客观规则联合辨证，实现了离散属性测度、相对属性权和合成属性权的计算方法。结果：构建了中医证候诊断系统，针对肝硬化样本病例学习辨证规则，并识别新样本的证候。结论：通过中医肝硬化数据进行系统的实例验证研究，表明该系统所得辨证结果与临床诊断吻合较好。

Syndrome Differentiation of Traditional Chinese Medicine Based on Attribute Hierarchical Model

Objective： To investigate a traditional Chinese medicine syndrome diagnosis model with subjective rule and objective rules. Methods： Based on the hierarchical model, mixed feature of traditional Chinese medicine with feature of western medicine, the decision table, integrated the subjective experience with objective machine-learning rules was construct to identify syndrome, implement the computation method about discrete attribute-measure, related attribute-weight and synthetic attribute-weight. Results： The system of syndrome differentiation was constructed, which study the syndrome differentiation rule about liver cirrhosis, diagnosis the syndrome of a new sample. Conclusion： The system is applied to identify the syndrome of liver cirrhosis,the experimental results are consistent with the clinical diagnosis.