earticle

영어

On demand resource forecasting in cloud computing is an crucial guarantee for achieving effective management of all virtualized resources and reducing data center energy consumption. According to single forecasting model cannot integrate all the valid information which leads to the decline in prediction accuracy. This paper proposed an optimal combination prediction model for cloud computing resource requirement. This model is based on generalized Dice coefficient and the induced ordered weighted geometric mean (IOWGA) operator, as well as improved Elman neural network and grey forecasting model. It is able to accurately reflect the random information and trend information in cloud computing load thus will enhance the overall prediction accuracy. The experiment results show this method is feasible and effective.