摘要
随着大型零售商企业信息化程度的不断提高,数据资源的深度挖掘和综合利用的要求日益迫切。FP-growth算法是在数据挖掘中应用最为广泛的一种关联规则挖掘算法,但存在占用内存大、运行速率慢和影响大数据价值发掘时效性等问题。提出了一种改进的FP-growth算法——MFP-tree算法,在继承FP-growth算法优点的前提下,采用分块挖掘的方式对数据进行挖掘,可有效提高效率。
As the informatization degree of big retailers is enhanced continuously, the deep mining and comprehensive utilization of data resource become more and more exigent. The FP-growth algorithm is the most widely used association rule mining algorithm in data mining, but it occupies more memory and runs very slowly, these disadvantages influences the mining timeliness of big data value. Auming at this problem, this paper puts forward MFP-tree algorithm which is improved based on FP-growth. The MFP-tree algorithm has still the advantages of FP-growth algorithm, and it adopts block mining method to improve the efficiency.
出处
《计算机与网络》
2017年第14期68-71,共4页
Computer & Network