Syndrome discrimination model of traditional Chinese medicine for chronic hepatitis B

Traditional Chinese medicine (TCM) has been widely applied in chronic hepatitis B (CHB), and syndrome discrimination is the most important step of TCM and it is performed subjectively and generally by physicians at present, which hinders the application prospects of TCM. In this paper, a CHB discrimination model of TCM is proposed basing on information gain, logistic attribute selection and space vector, through the approach, critical weighting attributes are selected, and cases are discriminated. The discriminant model is evaluated by CHB dataset, 34 critical weighting attributes are selected and 555 typical cases of two syndromes are identified respectively. And the selected critical attributes are in sound agreement with those used in TCM syndrome differentiation by physicians. Finally, results of this discriminant model for CHB are compared with other methods, and experimental results also show the discriminant model performs well in the application for TCM syndrome differentiation of CHB.