Research Of Association Rules Algorithm Based On Fuzzy Theory

To solve the problem in mining quantitative data with association rules, it is an important method that converts quantitative attributes into Boolean attributes by dividing its values into several sectors. In this dissertation, firstly, a improved Apriori algorithm which generates rules after scan discrete data set only once, was presented to enhance efficiency of mining Boolean data with association rules. Due to some problems of dividing values into sectors, such as boundary between sectors would be rigid, two solutions based on fuzzy theory were given. One is founded on fuzzy cluster matrixes for fixed sector number, another one is founded on fuzzy equivalent matrixes for unfixed sector number. And both of them have been bewritten, interpreted, analyzed in detail. Besides the above work, other discussion is how to calculate the support and confidence value based on degree of membership. In the last part of the dissertation, we applied these solutions as mentioned above to a monitoring system to make some analysis, which can well verify the effectiveness and usability.