Fig. 3
From: Efficient resource provisioning for elastic Cloud services based on machine learning techniques
Summary of input training data. This graph displays the input data used for the training of the SVM-based forecasting model presented in this work. This data were obtained from real web service logs from the Complutense University of Madrid and were gathered over a four-week period on an hourly basis. To emulate a server with high data traffic load, the data collected have been extrapolated one order of magnitude. These input training data represent the hourly average load (expressed in requests per second) of the server for a 4-week period (i.e., N=672 hours)