Date 9/800: Mining structure and model
Date 9/800: Mining structure and model
Structures and Models
As detailed in Chapter 3, SQL Server Analysis Services has two major objects
that deal with data mining: mining structures and mining models.
A mining structure defines the domain of a mining problem, whereas a
mining model is the application of a mining algorithm to the data in a mining
structure. A mining structure contains a list of structure columns that have data
and content types, bindings to the data source, and some optional flags that
control how the data is modeled. Additionally, a mining structure contains a
list of mining models that use the columns from the structure.
The definition of a mining model contains an algorithm with its associated
parameters, plus a list of columns from the mining structure. Each model in a
structure can use a different algorithm, or the same algorithm with different
parameters, and/or a different subset of the columns in the structure. For each
column in the model, you can assign how it is to be used in that model, as
well as algorithm-specific modeling flags. This feature allows you to easily test
different hypotheses on the same data set.
Structures and Models
As detailed in Chapter 3, SQL Server Analysis Services has two major objects
that deal with data mining: mining structures and mining models.
A mining structure defines the domain of a mining problem, whereas a
mining model is the application of a mining algorithm to the data in a mining
structure. A mining structure contains a list of structure columns that have data
and content types, bindings to the data source, and some optional flags that
control how the data is modeled. Additionally, a mining structure contains a
list of mining models that use the columns from the structure.
The definition of a mining model contains an algorithm with its associated
parameters, plus a list of columns from the mining structure. Each model in a
structure can use a different algorithm, or the same algorithm with different
parameters, and/or a different subset of the columns in the structure. For each
column in the model, you can assign how it is to be used in that model, as
well as algorithm-specific modeling flags. This feature allows you to easily test
different hypotheses on the same data set.
Date 9/800: Mining structure and model
