Research On The Improvement Of Association Rules Mining Algorithm And Its Application

Data mining is the process of discovering interesting knowledge from massive data which is stored in databases, data warehouses, etc. Data mining contains various techniques, and association rules mining is one of the most important data mining methods.Based on the analysis of association rules mining algorithm, this paper puts forward an improved algorithm for algorithm Apriori. Also, according to the characteristics of missing data, this paper applies the improved algorithm to the field of imputing missing data, thus a new multiple imputation method which is based on association rules is proposed. The main contents of this paper are listed as follows:The research on the improvement of association rules mining algorithm. Based on the introduction of data mining and association rules. Aiming at the shortage of Apriori algorithm, an improved algorithm TBapriori is proposed, which is based on the method of Binary Search. Through Comparative experiments among the TBapriori algorithm, the Apriori algorithm and other algorithms, the results show that the TBapriori algorithm has a significant reduction in the number of the candidate sets and its scanning times in the database reduces obviously.The research on the application of TBapriori algorithm. Based on analyzing and summarizing the characteristics of missing data, combining with the improved algorithm TBapriori, this paper presents a new multiple imputation method named MDMITB which can be used in imputing missing data. The method includes using TBapriori algorithm to generate strong association rules, the algorithm of sorting rule groups, multiple imputation, statistics and analysis. Finally, comparing the imputation accuracy of the new imputation method with MIKNN algorithm and MIRandom algorithm though experiments, the results show that the imputation accuracy of MDMITB method is higher than that of the others.