Multi-step-ahead Host Load Prediction with VMD-BiGRU-ED in Cloud Computing

Cloud computing resource load has the characteristics of nonlinearity, non-smoothness, high noise and high fluctuation, and it is difficult to achieve better prediction results by using a separate prediction model. Therefore, in this paper, we propose a combined VMD-BiGRU-ED prediction model based on VMD algorithm, GRU and encoder-decoder structure, which combines the advantages of strong VMD signal decomposition ability and high GRU prediction accuracy to accurately predict cloud computing resource load. To evaluate our model, we conduct experiments using the cluster public dataset provided by Alibaba Cluster-trace-v2018. The experimental results show that our proposed combined VMD-BiGRU-ED prediction model outperforms the single prediction model and other combined prediction models in terms of prediction accuracy.