Study on Applying Multiple Combinations to Network Traffic Analysis
Guo-xin Cao · Journal of Chinese Computer Systems · 2010
The conventional stable network traffic model is incapable of describing unstable characteristics.In the period of forecasting,the step increases while accuracy falls.Although NN (neural network) by way of an unstable model could make up for the defect of stable model,there has not been a theory to select parameters of the network yet.According to the problems above,a method of multiple combinations is presented.First,comparing ARIMA with Elman model then apply the multiple combinations to network traffic analysis.The experiment results show that the method is superior to a single model.