Short-Term Time Series Modelling Forecasting Using Genetic Algorithm

Abstract

The prediction analysis of a network traffic time series dataset in order to obtain a reliable forecast is a very important task to any organizations. A time AQ1 series data can be defined as an ordered sequence of values of a variable at equally spaced time intervals. By analyzing these time series data, one will be able to obtain an understanding of the underlying forces and structure that produced the observed data and apply this knowledge in modelling for forecasting and monitoring. The techniques used to analyze time-series data can be categorized into statistical and machine learning techniques. It is easy to apply a statistical technique [e.g., Autoregressive Integrated Moving Average (ARIMA)] in order to analyze time-series data. However, applying a genetic algorithm in learning a time series dataset is not an easy and straightforward task. This paper outlines and presents the development of genetic algorithms (GA) that are used for analyzing and predicting short-term network traffic datasets. In this development, the mean squared error (MSE) is taken and computed as the fitness function of the proposed GA based prediction task. The results obtained will be compared with the performance of one of the statistical techniques called ARIMA. This paper is concluded by recommending some future works that can be applied in order to improve the prediction accuracy.