Using the Data Mining Add-In for Microsoft Office - Microsoft Business Intelligence Tools for Excel Analysts - page 273

Following is a list of prerequisites for using the Forecast tool:
➤ One of the columns must be time-oriented or a unique ordered set of numbers (which could
substitute for an actual time/date column).
➤ The predictable column must be a continuous value. Bear in mind you will not be predicting
values for the time/date column.
➤ Before the model is generated, you can provide a hint to specify that your time-oriented data
has a cyclical pattern, as would be the case for monthly sales, for example.
The Forecast tool helps you make predictions based on data in an Excel data table or other data
source, and optionally view the probabilities associated with each predicted value. For example, if
your data contains a date column and a column that shows total sales for each day of the month, you
could predict the sales for future days. You can also specify the number of predictions to make. For
example, you can predict five days, or thirty.
To see the Forecast tool in action, click the Forecast button on the Analyze tab. In the Forecast dialog
box, shown in Figure 14-9, identify the following inputs: Columns to be forecast, Number of time
units to forecast, Time stamp (optional), and Periodicity of data (optional). Click the Run button and
you get a report similar to the one shown in Figure 14-10.
Figure 14-9: Configure the Forecast dialog box as needed.