A statistical research on feed forward neural networks for forecasting time series,

Abstract

In recent years, artificial neural networks have being successfully used in time series analysis. Using linear methods such as ARIMA and exponential smoothing for non linear time series cannot produce satisfactory results. Although there are various non linear methods, these methods have an important drawback that all of them require a specific model assumption. On the other hand, artificial neural networks have no restrictions such as linearity or model assumptions. In many applications within the time series analysis, it has been seen that artificial neural networks produce more accurate results than those obtained from traditional methods. In spite of the fact that artificial neural networks provide some advantages, researchers keep working on the component selection problem of the method. The answer of the question that which components of the method should be used is a vital issue in terms of forecasting performance. In this study, the effects of number of hidden layer and length of test set on forecasting performance of artificial neural networks are examined. Eight real time series are used in the implementation. The obtained results are analyzed by using statistical analysis and are interpreted.