Improving Time Series' Forecast Errors by Using Recurrent Neural Networks

Elman Neural Network (ENN) is considered one of the most powerful tool in solving various models. This paper suggests the use of ENN in a model free technique to solve time series models of any type. The objective of this paper is to compare between the suggested smart method against the traditional method in solving time series problems. The accuracy of the prediction method measures used in this paper are Root-Mean-Square Error (RMSE) and Mean Absolute Percentage Error (MAPE), thus evaluating the adopted prediction methods. The results show that suggested smart method which uses the ENN is better compared to the traditional method, which uses Autoregressive Integrated Moving Average (ARIMA) model to solve time series forecasting models.