The Research Of Application On Medical Data Process And Mining Algorithm Of Association Rules

With the fast development of the information technology, Data Mining is an emerging research and application topic. Association Rule Mining is an important branch and becomes one of the widest applied data mining styles. Based on the research of the algorithm on mining association rule, the thesis makes research on the analysis of medical data and the algorithm improvements of mining medical association rules. The main achievements are as follows:1. The basic concepts and theories of Association Rules are described. The typical algorithms: Apriori and FP-growth of mining association rules are discussed and analyzed in merits and defects of these algorithms. The improved algorithm on Apriori is brought forward and programmed to execute.2. The Association Rule mining algorithm DMA and FMGMFI in distributed database system are described. Based on the algorithm DMA, the improved algorithm on DMA is presented. It needs less memory and communication cost. These algorithms are implemented.3. The characteristics of medical data are analyzed. An example about medical data is given to explain how to map medical data to a transaction format containing items.4. Because of problems existing in the application of association rules in medical data, an improved algorithm on FP-growth is introduced to solve them. By testing a medical image data, it is shown that the algorithm reduces the uninteresting rules and outperforms the algorithm FP-growth. Based on this, a model of association based self-adaptive mining classification rules is proposed.