Association Analysis And Application Of Production Process Based On Apriori Algorithm

With the advent of the era of industrial big data,a large amount of production data has been generated in industrial production.Mining potential association rules and industry knowledge from these production data can provide rich suggestions and help for decision-making control in the production process,which is also one of the important branches of data mining.Due to the characteristics of industrial production data,such as large amount of data,multi-dimensional,and large amount of numerical data,it is difficult to reduce the running time and space of association rules in data mining.However,using the traditional Apriori algorithm for association analysis will cause problems such as too long data mining time,candidate itemsets bursting,and the generated frequent itemsets occupying too much memory.To solve the above problems,this paper proposes two improved association rule algorithms,and applies the improved association rule algorithms to the tobacco production process,mining the potential association relationship of production data,providing a reference basis for the control and decision-making of tobacco production.First,to solve the problem of low efficiency in generating frequent itemsets caused by the explosion of candidate sets,an Apriori algorithm based on dynamic hash is proposed.By combining the data characteristics of the dataset,the algorithm applies the dynamic hash technology to improve the efficiency of frequent itemsets generation.The simulation results show that compared with the classical Apriori algorithm,the running time of the algorithm is reduced by more than 30%,and the efficiency of generating frequent itemsets is effectively improved.Secondly,aiming at the problem that different feature datasets have less redundant deletion of candidate sets in data mining,an Apriori algorithm based on transaction compression is proposed.The algorithm adopts the method of deleting candidate sets in advance to improve the redundant transactions of the database,so as to reduce the number of redundant candidate sets.The simulation results show that this algorithm can save more than 20% of the database memory.Finally,the improved algorithm is applied to the tobacco production association rules mining,and the tobacco production data is used as the data set for association rules mining.According to the results of the excavation,the feasible suggestions are put forward for the parameter control of tobacco factory production.