云计算下大数据分布式存储中冗余数据分配算法

Redundant Data Allocation Algorithm for Distributed Storage of Large Data in Cloud Computing

传统支持向量机算法分配冗余数据时,据分配算法,采用局部特征分析算法获取云计算下大数据分布式存储中冗余数据的重要特征,将该特征设为分类冗余数据的基本数据;使用一对一的分类方式将基本数据变换成二分类数据后,采用最优分类平面对二分类数据进行深仅能进行小样本数据中冗余数据的分配,不能解决大数据中的冗余数据间关联性低的问题,具有分配效率和准确率较低等弊端。提出云计算下大数据分布式存储中冗余数度分类,获取冗余数据片段;采用基于遗传算法对冗余数据片段的分配过程进行求解,构建分配策略和评估准则,获取最佳冗余数据分配策略,完成对云计算大数据分布式存储中冗余数据的最优分配。实验结果说明,所提算法对云计算下大数据分布式存储中的冗余数据实施分配时是具备精度以及效率高的优势。

When the traditional SVM algorithm distributes redundant data,it can only allocate redundant data in small sample data,and can not solve the problem of low correlation between redundant data in large data,which has the disadvantages of low allocation efficiency and low accuracy. A redundant data allocation algorithm for distributed storage of large data in cloud computing is proposed. The local feature analysis algorithm is used to obtain the important features of redundant data in distributed storage of large data in cloud computing. This feature is set as the basic data of classified redundant data. After the basic data is converted into two classified data by one to one classification method,the optimal classification is used to face two classified data for depth classification. The redundant data fragments are obtained; the allocation process of redundant data fragments based on genetic algorithm is solved,the allocation strategy and evaluation criteria are constructed to obtain the best redundant data allocation strategy,and the optimal allocation of redundant data in the large data distributed storage of cloud computing is completed.The experimental results show that the proposed algorithm has the advantage of accuracy and efficiency in the allocation of redundant data in the distributed storage of large data under cloud computing.