Improvement And Application Of Apriori Algorithm For Association Rules

In recent years,road traffic accidents have occurred more and more frequently,and the causes of traffic accidents have become more and more complex.Therefore,the cause analysis of traffic accidents has become a hot topic of academic research in recent years.With the further deepening of the research,it is understood that data mining technology is widely used in the traffic field,and the association rule Apriori algorithm is a more widely used algorithm in data mining technology,which is designed to find the relationship between transaction data items.Compared with other data mining algorithms,it can better mine the relationship between the factors causing traffic accidents.This paper mainly studies the Apriori algorithm of association rules and applies it to mining association rules caused by traffic accidents.The main work and achievements are as follows:1.An improved Aprioir algorithm for association rules is proposed.Aiming at the problem that Apriori algorithm needs to scan the database for many times,the database is stored in the form of matrix,and the transaction support of candidate set is obtained by using binary coding method,so that the entire algorithm only needs to scan the database once;Aiming at the problem that the algorithm is easy to generate a large number of candidate item sets in the process of generating frequent item sets,the number of candidate item sets is reduced by improving the pruning strategy when k-order frequent item sets generate k-order candidate item sets;At the same time,the conditions for the end of the algorithm iteration are improved to reduce unnecessary transaction scanning.Through comparative experiments,it is proved that the optimization algorithm solves the problem of scanning the database for many times and generating too many candidate item sets,and improves the operation efficiency of the algorithm.2.Realize the mining of traffic accident association rules based on multi value attribute association rules.Due to the multidimensional nature of traffic accident data,this paper describes the data uniformly by reference to data coding,and then uses the improved Apriori algorithm to mine the traffic accident data to realize the data mining of multi value attribute association rules.In the process of data mining for traffic accident data,in view of the problem that the algorithm is easy to generate the connection of different values of the same attribute when searching the maximum frequent item set,the problem that the algorithm is easy to generate the connection of different values of the same attribute is solved by adding the method of testing the same attribute in the improved algorithm;Aiming at the problem that some meaningless association rules may exist in the generated association rules,the method of adding subjective constraints is used to filter the generated association rules,which provides a shortcut to find meaningful association rules;To solve the problem that the rules mined by only using the support and confidence measures may be invalid,KULC measure and balance ratio(IR)are added to determine the reliability of association rules.