Classification Based on Association with Atomic Association—Prediction of Tibetan Medicine in Common Plateau Diseases

Abstract

In order to make up for the deficiencies in the research of drug combinations in the study of the prescriptions of Tibetan medicines, this paper applies the Drug-atomic combination association rules to the study of Tibetan medicine prescriptions based on the intrinsic link between drugs. Based on the data of 223 chronic atrophic gastritis cases in Tibetan hospitals in Qinghai Province, using the correlation analysis and association rule algorithm, the method of using commonly used Tibetan medicine pairs for atomic combination to research was proposed for the first time; in the process of extracting common drug pairs, base on the method that limit the length of frequent item sets generated to improve the efficiency of the traditional Equivalence Class Clustering and bottom up Lattice Traversal algorithm (ECLAT algorithm). In this paper, the Classification Based on Association algorithm (CBA algorithm) based on the atomic combination of drugs is creatively applied to the prediction of Tibetan medical drugs. Base on the pulse, urine, and tongue diagnosis data of chronic atrophic gastritis with Tibetan characteristics to predict the diagnostic medicine of doctors. Used the class association rules obtained before were applied to chronic atrophic gastritis in Tibetan medical treatment data, and a 5-fold cross-validation of the rule set was performed. The accuracy rate was 76%. This achieved the preliminary goal of predicting Tibetan medicines based on Tibetan medical vein, urine, and tongue diagnostic data.