基于改进的包容性检验的云计算资源组合预测  

Combined forecasting of cloud computing resources on improved encompassing tests

作　　者：李洪安[1] 康宝生[1] 张婧[1] 杜卓明[1] 朱煜[1] 

机构地区：[1]西北大学信息科学与技术学院,西安710127

出　　处：《计算机应用研究》2013年第1期252-255,共4页Application Research of Computers

基　　金：西北大学研究生创新教育项目(10YSY02);陕西省教育厅自然科学专项研究资助项目(11JK1053)

摘　　要：针对云平台资源的组合预测中包容性检验计算量大且准确性较低的问题,提出了一种改进的包容性检验方法并应用于云平台资源的组合预测中。通过按单一预测模型的优劣次序逐步进行包容性检验并与主成分分析组合预测相结合,以此思路形成了改进的包容性检验方法。仿真实验结果表明,改进的包容性检验方法不仅在计算效率及精度上优于现有的包容性检验方法,而且提高了云平台资源的组合预测的性能。To solve the problems that the encompassing tests in combined forecasting of cloud computing resources have a high computation and a low accuracy,this paper proposed and applied a method of improved encompassing tests to the combined forecasting of cloud computing resources.It made the encompassing tests gradually in the order of the single forecasting model＇s relative merits,and with the combination of forecasting based on principal component analysis,which formed the method of improved encompassing tests.The simulation results show that the method of improved encompassing tests can not only increase the calculation efficiency and accuracy,but also improve the performance of the combined forecasting of cloud computing resources.

关 键 词：云计算 包容性检验 主成分分析 组合预测模型 

分 类 号：TP393.06[自动化与计算机技术—计算机应用技术]