SAS/ETS Software

Automatic Forecasting

SAS/ETS software includes many procedures that produce forecasts and confidence limits.

Plot of forecasted values from the UCM Procedure
Plot of forecasted values from the UCM Procedure

The AUTOREG Procedure

The AUTOREG procedure can be used to model and forecast time series data when your data can be fitted or approximated by an autoregressive model. The forecasts produced by PROC AUTOREG are dependent upon the type of model and estimation method used to fit the data.

Details of the AUTOREG Procedure.

The ARIMA Procedure

The ARIMA procedure utilizes the Box-Jenkins approach to time series data. Like PROC AUTOREG, the ARIMA procedure produces forecasts for autoregressive models. One advantage of PROC ARIMA is its ability to identify the error process underlying the time series. Additionally, PROC ARIMA enables you to fit more complicated models. The forecasts produced by the ARIMA procedure depend upon the type of model fit and the estimation method used to fit the model. In some cases, it may be more efficient to use PROC ARIMA instead of PROC AUTOREG to identify models of unknown processes.

Details of the ARIMA Procedure.

The ESM Procedure

The ESM procedure provides quick and simple forecasting of one or more time series by exponential smoothing. It is useful when you want to generate quick forecasts of multiple time series without having to develop a model for each series. The ARIMA and AUTOREG procedures may produce more accurate forecasts and confidence intervals, but they require more time and input from you.

Details of the ESM Procedure.