摘要
论文针对改进后的关联规则挖掘算法--基于矩阵的Apriori算法所存在的不足,在此算法的基础上对其进行了进一步的改进,给出了一种基于压缩矩阵的高效关联规则挖掘方法CMEAR算法。该方法通过对矩阵的压缩,改进矩阵的存储方式及对项目集进行排序等多种方式实现关联规则的挖掘。最后通过实验将该方法与传统Apriori算法以及基于矩阵的Apriori算法进行了对比分析,实验结果表明基于压缩矩阵的高效关联规则挖掘方法使算法在时间性能上有很大的提高。
Aiming at the shortcomings of Apriori algorithm based on matrix for the improved association rules mining algorithm,further improvement is made based on this algorithm,a kind of efficient association rule mining algorithm based on compressed matrix CMEAR is proposed in this paper.This method realizes the association rules mining by compression of matrix,improvement of storage mode of matrix and sorting of item set.Finally,the improved method is compared with traditional Apriori algorithm and Apriori algorithm based on matrix.The experiment results proves that the efficient association rule mining algorithm based on compressed matrix makes a great improvement in time performance.
作者
潘俊辉
张强
王辉
王浩畅
PAN Junhui;ZHANG Qiang;WANG Hui;WANG Haochang(Northeast Petroleum University,Daqing 163318)
出处
《计算机与数字工程》
2019年第11期2819-2823,共5页
Computer & Digital Engineering
基金
国家自然科学基金项目(编号:61702093)
黑龙江省教育厅科研专项——东北石油大学引导性创新基金项目(编号:2016YDL-12)资助
