A hybrid method for forecasting trend and seasonal time series

Forecasting of time series that have trend and seasonal variations remains an important problem for forecasters. In this work, a hybrid method which combines Winters´ exponential smoothing method and neural network is proposed for forecasting seasonal and trend time series. The proposed method aims to integrate the linear characteristics of an exponential smoothing model and nonlinear characteristics of neural network to create a more effective model for time series forecasting. Experimental results show that the hybrid method outperforms neural network model in forecasting seasonal and trend time series.