Variable smoothing parameter of the double exponential smoothing forecasting model and its application

The exponential smoothing method is widely used in time series forecasting. However, the smooth parameter α in single exponential smoothing model and double exponential smoothing model is unchanged in conventional practice. This paper has improved the conventional exponential smoothing method, at the second exponential smoothing we introduce another smoothing parameter β. We model the variable smoothing parameter of the double exponential smoothing forecasting model as MSE or MAD for the optimization objective, use autocorrelation parameter T and χ² method is superior to the conventional method.