7. UNIVARIATE MODEL FORECASTING


The Univariate Model Forecasting software allows you to
investigate the behavior of historical data series and
forecast future observations with a univariate model.  This
facility is designed to address three requirements:

o  Fitting linear, quadratic, and cubic models to historical
   data series.

o  Smoothing historical data series to minimize problems
   introduced by random variations in the historical data
   observations.

o  Allowing you to modify and/or delete historical data
   points to deal with errant data.

Note: SAS/STAT is a prerequisite for using Univariate Model
Forecasting.

This section contains the following topics:

7.1 Functional Description

7.2 Usage Guidelines

7.3 Standard Output

7.4 Analytic Technique Tutorial

7.5 Component Operation

7.6 Case Studies