SAS Forecast Studio is the client component of the SAS Forecast Server,
which is an example of the SAS analytics products that use the SAS
Intelligence Platform. The SAS Forecast Server is a large-scale, automatic
forecasting solution that enables organizations to produce huge quantities
of high-quality forecasts quickly and automatically. SAS Forecast
Studio provides a graphical interface (based on Java) to the forecasting
and time series analysis procedures contained in SAS High-Performance
Forecasting and
SAS/ETS software.
SAS High-Performance
Forecasting automatically selects the appropriate model for each item
being forecast, based on user-defined criteria. Holdout samples can
be specified so that models are selected not only by how well they
fit past data but by how appropriate they are for predicting the future.
If the best forecasting model for each item is unknown or the models
are outdated, a maximum level of automation can be chosen in which
all three forecasting steps (model selection, parameter estimation,
and forecast generation) are performed. If suitable models have been
determined, you can keep the current models and reestimate the model
parameters and generate forecasts. For maximum processing speed, you
can keep previously selected models and parameters and choose to simply
generate the forecasts.
For more information,
see the
SAS High-Performance Forecasting: User's Guide and the
SAS/ETS User's Guide.