Research And Application Of Association Rules Mining Based On Hybrid Rice Optimization Algorithm

With the rapid development of Internet technology,the application of data mining technology has become increasingly widespread,and people's requirements for data mining accuracy and efficiency are also increasing.However,in the face of increasing data volume,many traditional algorithms cannot solve the data quickly and effectively.Therefore,the application of new evolutionary algorithm in data mining has received extensive attention.Hybrid Rice Optimization Algorithm(HRO)is a new type of swarm intelligence algorithm,which has many advantages such as low paramete and strong global optimization ability,as well as the advantage that it is not easy to fall into local optimal solution.In order to adapt to the processing of large data sets,the hybrid rice optimization algorithm is applied to association rule mining,and the hybrid rice optimization algorithm is parallelized.The specific work is as follows:1.The research of association rule mining based on hybrid rice optimization algorithm is conducted.On account of the fact that Apriori algorithm,as the classic association rules mining algorithm,has the disadvantages of long time consumption,large memory consumption,and easy extraction of useless rules,this paper proposes the mining of association rules based on hybrid rice optimization algorithm,and adopts real-coded method to set rice,and selects reasonable fitness function.And binary coding was used to mine the optimal rules from the data set.In the process of experimental simulation,a comparison with the particle swarm algorithm,genetic algorithm and Apriori algorithm was done.It proved that the hybrid rice optimization algorithm was effective in mining association rules.2.Parallel Hybrid Rice Optimization(PHRO)was studied.The hybrid rice optimization algorithm was run on the Hadoop platform,and the parallelized hybrid rice optimization algorithm was implemented by using the MapReduce computation framework to improve the convergence speed and computation speed of the algorithm.Through experimental simulation,parallelized hybrid rice optimization algorithm was compared with traditional serialized hybrid rice optimization algorithm in association rule mining,and IBM public data set was used to compare the algorithm performance of them.The results show that association rules mining based on the parallelized hybrid rice optimization algorithm is significantly serialized in the calculation of the association rules of the hybrid rice optimization algorithm.In general,the hybrid rice optimization algorithm was applied to association rule mining.The experimental results show that the hybrid rice optimization algorithm has good performance in mining association rules and has a good application prospect.