Analysis Based On The Prescriptions Of The Imprecise Reasoning

Traditional Chinese Medicine (TCM) prescription embraces abundant knowledge and experience, thus the TCM prescription experience mining consists paramount importance in TCM datamining. From prescription effect and composing these two perspectives, this paper adopts kinds of datamining methods to study TCM prescription. First of all, prior knowledge is obtained through mathematics and DM methods, then with every single prescription set as the research object, staring from medication and dosage, varieties of DM and fuzzy mathematics methods are used to analyze and summarize the prescription effects, dose-effect relations of couplet medicines, the corresponding syndromes and therapies so as to reflect the TCM treatment with syndromes differentiation thinking. On the other hand, prescription composing is studied with fuzzy thinking, in this way not only can we better understand the prescription composing rules but also certain theory proofs for new prescription composing trials can be provided. Specific work is described as follows:1. TCM data pre-processing is studied, herbal medicine names, effects, symptoms and syndromes are formatted based on standards, and medicine dosages are transformed from actual dosages to relative dosages, known as the dosage intensity in this paper.2. Inter-effect mutually influencing relations are studied with association rules analysis, and relations between effects and symptoms are explored through SimRank, thereby "inter-effect" and "symptom-effect" relations are organized.3. On the basis of "inter-effect" relation, certain mathematics module is designed to study the prescription effect quantization, and "symptom-effect" relation is referenced to select and summarize the effective efficacies, leading to the achieving of "dosage-effect" relation.4. Bayesian method is used to study the "symptom-syndrome" relation, which is inferring syndromes through a group of symptoms, and corresponding therapies are fetched.5. Clustering and fuzzy association rule methods are integrated to dig and analyze the dose-effect relations of couplet medicines.6. TCM prescription composing and compatibility rules are studied, which is implemented with perturbation thinking based fuzzy clustering method and fuzzy intention decision based method.