A SVR Model for Differentiation of Syndromes in Traditional Mongolian Medicine

Abstract

Traditional Mongolian medicine (TMM) is the unique medicine of Mongolia nationality. It is meaningful to research, protect and develop TMM. In TMM, there are different causes for a disease. A doctor of TMM must not only diagnose what disease that a patient has but also what syndrome of the disease. Differentiation of syndromes (DS) in TMM is to distinguish diseases and their syndromes. DS is the key and difficult point of medical diagnosis and treatment. Doctors mainly do it according their experiences. So, it is necessary and useful to establish aided diagnosis models of DS for TMM. This study aims to establish a Support Vector Regression (SVR) model of DS. Firstly, we deduced the SVR. Then, we used medical cases to train the SVR model. Finally, we did experiments of validation. The experiments shown its correct rate was 76.6%.