Data Association Rules Mining Method Based on RBF Neural Network Optimization Algorithm

In order to achieve real-time and accurate data mining, this paper proposes a data association rules mining method based on RBF neural network optimization algorithm. On the basis of considering the constraint association rules, the data frequent itemsets are reduced to get the corresponding candidate data sets that meet the rules, and the update of the data frequent itemsets for user needs is completed. The candidate data set is input to RBF neural network for training, the network output is optimized by combining with rough set theory, the user demand data table and quadruple are constructed, the user demand attributes are described, and the user demand data is obtained by using upper bound pruning method. Through the upper bound pruning method, the user demand data is obtained and real-time mining is realized. Compared with other data mining algorithms, the simulation results show that the performance of the data mining algorithm proposed in this paper is less constrained by the proportion of redundant data and the size of data, the accuracy of data mining is higher, and it has better stability.