Classification Method for Interval Valued Relational Database.

BW Xu,JJ Lu,WC Chu,HJ Yan

2003-01-01

Abstract:Classification and association rules are important research issues and focuses of data mining technology. A classification method of association rules with linguistic variables, which fits for the interval valued relational database, is presented In this classification method, values of the record in each attribute are partitioned into several linguistic variables by the fuzzy c-medoids algorithm. Then, Apriori algorithm is improved for mining interesting association rules with linguistic variables in the interval valued relational database. Last, we use these interesting association rules with linguistic variables to build classification system. The classification method is implemented on a constructed interval valued relational database. The experiment results show that the classification method has fine accuracy.