邹华


摘 要: 针对云计算环境下大数据分布不合理,导致运行效率差的问题,提出基于模糊聚类分析的云计算环境下大数据分布规律的结构优化方法,首先对特征的观测值进行规格化处理,保证每个特征值的取值均处于[0,1]范围内。其次获得在[0,1]范围内的相似系数描述数据间的相似度,通过相似矩阵获得云计算环境下大数据集上的模糊五阶相似矩阵。引入最短距离法,将其和相似矩阵融合,共同实现数据聚类,完成大数据分布规律结构的优化设计。仿真实验结果表明,所提方法使得数据分布更合理,而且运行效率和稳定性均较高。
关键词: 云计算; 大数据; 分布规律; 相似矩阵
中图分类号: TN911?34; TP393 文献标识码: A 文章编号: 1004?373X(2016)08?0018?03
Optimization design for structure of big data distribution regularity
in cloud computing environment
ZOU Hua
(Information Engineering College, TongRen University, Tongren 554300, China)
Abstract: Aiming at the poor operating efficiency caused by the unreasonable distribution of big data in cloud computing environment, a structure optimization method of big data distribution regularities in cloud computing environment, which based on the fuzzy cluster analysis, is proposed. Firstly, a normalization treatment for the observed values of the features is performed to ensure that the values are all controlled in the range of [0,1]. And then the similarity factor in the range of [0,1] is acquired to describe the similarity between the data, and the fuzzy five?order similar matrix on big data set in cloud computing environment is obtained according to the similar matrix. The single linkage method is used to integrate with the similarity matrix to achieve data clustering, and realize the optimization design of large data distribution structure. The simulation results show that this method not only can make the data distribution reasonable, but also has advantages of high operating efficiency and high stability.
Keywords: cloud computing; big data; distribution law; similar matrix
随着计算机科技技术的逐渐发展,所涉及的数据量越来越大,人们对云计算环境安全性的要求也越来越高[1?2]。如何在不增加成本的情况下,提高整个系统的安全性和数据分布的合理性已成为相关学者研究的重点课题,受到了越来越广泛的关注[3?5]。目前,研究云计算环境下大数据分布规律的结构优化设计方法有很多,主要包括模糊聚类方法、详细度量方法和最小二乘法等,相关研究也取得了一定的成果。其中,文献[6]提出基于误差和曲线分析的云计算环境下大数据分布规律结构优化设计方法,通过误差分析法和实测数据直方图进行比较实现结构优化,但该方法存在实践过程复杂的弊端。文献[7]提出基于非线性最小二乘法的云计算环境下大数据分布规律结构优化设计方法,通过非线性最小二乘拟合法,获取数据的统计分布规律,但该方法的计算结果精度低,同时受外界环境的干扰较大。……