Método GetAttributes
Gets a collection of attributes based on their feature selection status.
Namespace: Microsoft.AnalysisServices.AdomdClient
Assembly: Microsoft.AnalysisServices.AdomdClient (em Microsoft.AnalysisServices.AdomdClient.dll)
Sintaxe
'Declaração
Public Function GetAttributes ( _
filter As MiningFeatureSelection _
) As MiningAttributeCollection
'Uso
Dim instance As MiningModel
Dim filter As MiningFeatureSelection
Dim returnValue As MiningAttributeCollection
returnValue = instance.GetAttributes(filter)
public MiningAttributeCollection GetAttributes(
MiningFeatureSelection filter
)
public:
MiningAttributeCollection^ GetAttributes(
MiningFeatureSelection filter
)
member GetAttributes :
filter:MiningFeatureSelection -> MiningAttributeCollection
public function GetAttributes(
filter : MiningFeatureSelection
) : MiningAttributeCollection
Parâmetros
- filter
Tipo: Microsoft.AnalysisServices.AdomdClient. . :: . .MiningFeatureSelection
The MiningFeatureSelection value to filter on.
Valor de retorno
Tipo: Microsoft.AnalysisServices.AdomdClient. . :: . .MiningAttributeCollection
A MiningAttributeCollection containing the attributes that have the specified feature selection value.
Comentários
When you create a mining model, Analysis Services chooses a subset of attributes to include in the data mining model based on statistical analysis. This attribute reduction process is called feature selection. The method for feature selection differs depending on the algorithm. For more information, see Seleção de recursos em mineração de dados.
Feature selection can be applied to both input and output attributes. If you have many attributes, in a nested table, for example, it is possible that not all of the attributes will be included in the final model. Therefore, you can call the MiningModel method to determine which attributes are selected for input only, which attributes are selected for output only, and which attributes are selected for both output and input. You can also return all attributes or attributes that have no feature selection value.