Research on the Hypertension Syndrome Elements Differentiation of TCM Based on Multi-label Learning and Ensemble Learning
Jun Li, Jianqiang Li, Dongyi Yi · 2018
Hypertension is one of the important causes of cardiovascular diseases. In China, the prevalence of hypertension is showing an upward trend. Syndrome elements differentiation has been given great attention in the field of Traditional Chinese Medicine (TCM) in recent years. This paper mainly focused on studying hypertension syndrome elements differentiation by multi-label learning and ensemble learning. Firstly, after comparing a several of algorithms' capability, a method of ECC+RF was proposed to build a classification model which showed a high performance. And then, MRUS algorithm was introduced with ECC+RF to handle the unbalanced problem about our data. The sample data consist of multiple symptoms and syndrome elements. It can be seen that our approach had a high performance and high interpretability via our numercial experiments. In addition, this work can bring the auxiliary decision and also it can find rules for the diagnosis of hypertension in TCM, which benefits the informative, objective and standardization in Chinese medicine.