| Forecasting Process Details |
Forecasting Process Summary
This section summarizes the forecasting process.
Parameter Estimation
The parameter estimation process for ARIMA and smoothing models
is described graphically in Figure 30.1.
Figure 30.1: Model Fitting Flow Diagram
The specification of smoothing and ARIMA models
is described in Chapter 25, "Specifying Forecasting Models,".
Computational details for these kinds of models
are provided in the following sections
"Smoothing Models" and "ARIMA Models."
The results of the parameter estimation process are
displayed in the Parameter Estimates table of the
Model Viewer window along with the estimate of the model variance
and the final smoothing state.
Model Evaluation
The model evaluation process is described graphically in Figure 30.2.
Figure 30.2: Model Evaluation Flow Diagram
Model evaluation is based on the one-step-ahead prediction errors
for observations within the period of evaluation.
The one-step-ahead predictions are generated from the model specification
and parameter estimates.
The predictions are inverse transformed (median or mean) and
adjustments are removed.
The prediction errors (the difference of the dependent series
and the predictions) are used to compute
the statistics of fit,
which are described in the following section
"Diagnostic Tests and Statistics of Fit."
The results generated by the evaluation process are
displayed in the Statistics of Fit table of the
Model Viewer window.
Forecasting
The forecasting generation process is described graphically in Figure 30.3.
Figure 30.3: Forecasting Flow Diagram
The forecasting process is similar to the model evaluation process
described in the preceding section,
except that k-step-ahead predictions are made from the end of the data
through the specified forecast horizon,
and prediction standard errors and confidence limits are calculated.
The forecasts and confidence limits are
displayed in the Forecast plot or table of the
Model Viewer window.
Copyright © 1999 by SAS Institute Inc., Cary, NC, USA. All rights reserved.
