Abstract—In this paper, a new hybrid model for predicting the exchange rate time series is introduced, using the multilayer perceptron (MLP) and radial basis function (RBF) neural networks to reduce the error of autoregressive integrated moving average (ARIMA) method. The hybrid model tries to detect the error of the linear statistical method and then model this error with MLP neural network. Again the remainder error is modeled with RBF neural network to reduce the final error of the hybrid model. In this two level process of error modeling, it will be proved that the final result of prediction and modeling is better than the results that could be a chieved by asingle ARIMA method or a single MLP or RBF neural network.
Index Terms—ARIMA; multilayer perceptrons; radial basis functions; time series forecasting
The authors are with Iran University of Science and Technology, School of Industrial Engineering, Iran (e-mail: firstname.lastname@example.org).
Cite: Arash Negahdari Kia, Mohammad Fathian, and M. R. Gholamian, "Using MLP and RBF Neural Networks to Improve the Prediction of Exchange Rate Time Series with ARIMA," International Journal of Information and Electronics Engineering vol. 2, no. 4, pp. 543-546, 2012.