For Tcm Stomach Diagnosis And Treatment Of Data Mining Technology

Chinese medicine has a long history and a remarkable curative effect. It is a science of studying the process of human life activities, resisting diseases and safeguarding the health .An old famous herbal doctor is an expert who formed in the long-term medical practice and has rich experience and wisdom. The summary of an old famous herbal doctor's experience is not only enriching the theoretical system of Chinese medicine, but also having a tremendous role in promoting the academic progress of Chinese medicine. In addition, the inheritance of an old famous herbalist doctor's experience is one of important ways which also brings up a new famous generation of herbal doctor.In order to inherit and develop the diagnosis knowledge of famous herbal doctors, this thesis uses TCM (Traditional Chinese Medicine) gastritis medical records as data samples, and through applying data mining algorithms analyze the relations between symptoms and diagnosis, and how to judge whether could get diagnosis according to symptoms.In this thesis, after pre-processing the Chinese medicine Gastritis Medical records, we use association rules methods to find the related relation between symptoms and diagnosis , diagnosis and recipe. Then we choose the symptoms which have strong related relation as candidate attributes, using decision tree algorithm constructing a decision tree according symptoms to classify. For some over-fitting exist in decision tree algorithm, the post-pruning algorithm is used to pruning the decision tree, optimizing the results. At last, Bayesian network is used to construct network topology between symptoms and diagnosis and reveal the relation of them. The experiment results show that using data mining related algorithms could find rule for TCM differential treatment, which interpretation the connotation and extension of old famous herbal doctors experience.