Time series forecasting using an ensemble model incorporating ARIMA and ANN based on combined objectives

Forecasting accuracy can be substantially improved through the combination of multiple individual forecasts. Yet most previous attempts at combining forecasts have focused on a single objective. This paper provides an ensemble forecasting model integrating autoregressive integrated moving average (ARIMA) with artificial neural networks (ANN) based on combined objectives. Golden section criteria is used in deciding the weight of two objectives. This method is examined by using the data of Canadian Lynx data series. Experimental results indicate that the combined model can be an effective way to improve forecasting accuracy achieved by either of the models used separately.