Evaluation of chaotic internet traffic predictor using MAPE accuracy measure

Abstract

The main aim of this article is to show that the presented prediction algorithm is accurate enough to perform good forecast of network traffic. The algorithm was presented in [1,2] as forecasting the stock market. This publication presents the results of appliance of this predictor for the internet traffic forecasting. It will be used in the network traffic prediction system so therefore the need of its exploration. The main contribution of the authors is exploration of the algorithm by measuring its accuracy using the MAPE (Mean Absolute Percentage Error) that is commonly used to assess the accuracy of prediction algorithms. It means gathering the data from the laboratory setup, elaboration of dedicated functions that control the algorithm and realization og the experiments. In the article we show only hypothetical results but essential. Our results show that the MAPE error of the predicted data is less than 10%, what gives very good result taking into account that MAPE between 20 and 30 percent means good prediction. © 2011 Springer-Verlag.