Discrete vs Continuous Variables

Some terms that constantly popup when learning to use the various data mining algorithms is discrete vs continuous variables. Many discussion center on whether the algorithm is suitable for predicted values or the input data are discrete or continuous.

Discrete variables take on fixed values. For example, values like True/False, integer values are discrete.

Continuous variables usually refer to real numbers (e.g. float, double, etc).
If there is a need for continuous data to be discretized, you can always define "bins" (denoting a specific range), and adding all those data that are within the range to the bin.

Another nice thingy about SQL Server 2005 is that while building your mining models, and you are unsure whether the input data should be discrete or continuous, the wizard will help you run through the data and recommend whether it is discrete or continuous (i.e. just use the wizards for a start!)