This article is a continuation of the following article that introduces the system stored procedures for SQL Server Data Mining.
 
http://social.technet.microsoft.com/wiki/contents/articles/7566.a-guide-to-the-undocumented-system-stored-procedures-for-data-mining.aspx

Neural Network and Logistic Regression Models

The following stored procedures are either used only with neural network and logistic regression models, or have special application to these models.

 

GetAttributeScores

Takes two sets of predictable values as inputs, and returns a table that shows the distribution of input values correlated to these outputs.

Fully qualified procedure name

System.Microsoft.AnalysisServices.System.DataMining.NeuralNet.GetAttributeScores

Parameters

Model name as string

Input definition as XML string

Attribute name as string

First attribute value as string

Valuetype as integer

Second attribute value as string

Valuetype21 as integer

Threshold as double

Parameters

The following example returns a list of the attributes in the specified model that are correlated to the outcome and value named in the second and third arguments.

CALL System.Microsoft.AnalysisServices.System.DataMining.NeuralNet.GetAttributeScores('Call Center NN', 'XML text, 'Service Grade', '0.090 - 0.115', 1, '0.115 - 0.139', 1, 0.0000005)

Parameters

The XML input definition always takes the following format.

<NewDataSet>   
<xs:schema id="NewDataSet"
xmlns=""
xmlns:xs="http://www.w3.org/2001/XMLSchema"
xmlns:msdata="urn:schemas-microsoft-com:xml-msdata">
  
<xs:element name="NewDataSet" msdata:IsDataSet="true" sdata:UseCurrentLocale="true">
  <xs:complexType>
     <xs:choice minOccurs="0&q=maxOccurs="unbounded">           
         <xs:element name="Table1">
            <xs:complexType>
               <xs:sequence>
                 <xs:element name="InputName" type="xs:string" minOccurs="0" />
                 <xs:element name="InputValue" type="xs:string" minOccurs="0" />
                 <xs:element name="InputValueType" type="xs:string" minOccurs="0" />
               </xs:sequence>
            </xs:complexType>
          </xs:element>
     </xs:choice>
  </xs:complexType>
</xs:element>
</xs:schema
</NewDataSet>

 

Sample results

The table shows the columns returned by the procedure call.

 

Attributes

Values

Scores

NodeProbability1

NodeProbability2

MarginalProbability1

MarginalProbability2

 

 

 

 

 

 

olid none;border-color:#000000 black black #000000;padding:0in 0.5pt;width:14.28%;background-color:transparent;">

 

 

 

 

 

 

 

 

 

 

 

Notes

[in progress] 

See also

No additional resources

Return to list of stored procedures for all model types