Fig. 6
From: Efficient resource provisioning for elastic Cloud services based on machine learning techniques
Accuracy of different forecasting models (real prediction error for the forecasting period) This figure shows the accuracy of the different forecasting models for the 24 h test interval (forecasting horizon), where the MAE, MSE, and RMSE values represent the real prediction error of each forecasting model for this period. They are computed as the error between the data obtained by the forecasting model and the real data of the time series for the forecasting horizon