A Method of Improvement and Optimization on Association Rules Apriori Algorithm

Abstract

The efficiency of mining association rules is an important field of knowledge discovery in databases. The algorithm a priori is a classical algorithm in mining association rules. A novel procedure was proposed to delete many transactions which need not be scanned repeatedly. The procedure described in this paper reduced the number of database passes to extract frequent item sets. A method was showed to reduce the number of candidate item sets by optimizing the join procedure of frequent item sets. To this end, the I a priori algorithm for mining frequent item sets, which is the improvement algorithm of a priori, is designed in this article. By a number of experiments, the proposed algorithm outperforms the a priori algorithm in computational time. The simulation results of knowledge acquisition for fault diagnosis also show the validity of I a priori algorithm