Association rules and decision tree based on traditional Chinese medicine diagnosis of gastritis.docVIP

 PAGE \* MERGEFORMAT 13 Association rules and decision tree based on traditional Chinese medicine diagnosis of gastritis [Keywords:] Data mining, association rules, decision trees, TCM Traditional use of “look, smell, and asked, cut” diagnostic approach, and to syndrome differentiation of chronic gastritis has a good effect. But as time goes by, some valuable resources and were not kept, the old Chinese medicine experience is the need of future generations on the basis of practice can only continue to comprehend and sum up the expression. need a lot of clinical data to extract valuable and common information to help determine, so as to achieve the purpose of syndrome differentiation. Therefore, I Using data mining association rules and decision trees in the approach to gastritis old TCM medical information for the object, the “symptoms”, “dialectical” relationship between the potential and under the “symptoms” to determine whether to get a “dialectical” one of doing. Based on the above purpose, the author uses the existing Traditional gastritis as the sample data records, using methods of association rules established experimental model, and gives experimental results of the analysis. On this basis, using the decision tree method, to determine whether building a dialectic “ virtual stagnation in the “decision tree. A method based on association rules of traditional Chinese medicine gastritis Data mining association rules is the most commonly used and proven methods. Association rule mining problem is given transaction database, to find out to meet the minimum support (minsup) and minimum confidence (minconf) of association rules. association rule has the following advantages: can produce useful results clearly support the indirect data mining, can handle variable-length data, the calculated consumption is expected [1]. classical association rule mining algorithm [2]: Apriori algorithm [ 3], sampling algorithm, DIC algorithm. About 1.1 Apriori