A Fuzzy Filter Based Hybrid ARIMA-ANN Model for Time Series Forecasting

This paper presents a new hybrid ARIMA-ANN model for time series forecasting. In this model, the time series is first decomposed into low-volatile and high-volatile components using a fuzzy filter. The low-volatile component is modeled using ARIMA and high-volatile component is modeled using ANN. The final prediction is obtained by combining the predictions from ARIMA and ANN models. Five real world time series datasets are used for comparative performance analysis of the proposed methodology with ARIMA, ANN and some existing hybrid ARIMA-ANN models. Experimental results show the superiority of proposed model than the other models considered.