Creating Data Mining Models
Use the Mining Model Wizard to create new data mining models. The Mining Model Wizard takes you through several steps needed to establish the model type, build the model the case set will work with, and choose the data mining technique the model will use to construct a new data mining model.
There are two types of data mining models used, based on the type of case set data to be processed. A relational data mining model is designed to process traditional relational database tables, while an OLAP data model is designed to process OLAP data stored in the form of cubes. The creation of each type of mining model is covered in its own topic.
Data mining models share the same schemas and basic structures, regardless of whether they are based on relational or OLAP data. The most important element in the construction of a data mining model is the case. The case is a set of data mining columns used to define the information that the model will use to identify and study, and for which the model will provide prediction data.
Data mining columns vary in content type: Key columns are used to identify a specific case, input columns are used to provide information that the data mining provider can use to analyze a case, and predictive columns provide meaningful results, in the form of histogram data, from the analyzed case.
See Also
Creating OLAP Data Mining Models