Düzenle

Aracılığıyla paylaş


Data Mining Model Training Destination

Applies to: SQL Server SSIS Integration Runtime in Azure Data Factory

Important

Data mining was deprecated in SQL Server 2017 (14.x) Analysis Services and now discontinued in SQL Server 2022 (16.x) Analysis Services. Documentation is not updated for deprecated and discontinued features. To learn more, see Analysis Services backward compatibility.

The Data Mining Model Training destination trains data mining models by passing the data that the destination receives through the data mining model algorithms. Multiple data mining models can be trained by one destination if the models are built on the same data mining structure. For more information, see Mining Structure Columns and Mining Model Columns.

Configuration of the Data Mining Model Training Destination

If a case level column of the target structure and the models built on the structure has the content type KEY TIME or KEY SEQUENCE, the input data must be sorted on that column. For example, models built using the Microsoft Time Series algorithm use the content type KEY TIME. If input data is not sorted, the processing of the model may fail. If the data requires sorting, you can use a Sort transformation earlier in the data flow to sort the data. This requirement does not apply to columns with the KEY content type. For more information, see Content Types (Data Mining) and Sort Transformation.

Note

The input to the Data Mining Model training destination must be sorted. To sort the data, you can include a Sort destination upstream from the Data Mining Model Training destination in the data flow. For more information, see Sort Transformation.

This destination has one input and no output.

The Data Mining Model Training destination uses an SQL Server Analysis Services connection manager to connect to the Analysis Services project or the instance of Analysis Services that contains the mining structure and mining models that the destination trains. For more information, see Analysis Services Connection Manager.

You can set properties through SSIS Designer or programmatically.

The Advanced Editor dialog box reflects the properties that can be set programmatically. For more information about the properties that you can set in the Advanced Editor dialog box or programmatically, click one of the following topics:

For more information about how to set properties, see Set the Properties of a Data Flow Component.

Data Mining Model Training Editor (Connection Tab)

Use the Connection page of the Data Mining Model Training Editor dialog box to select a mining model to train.

Options

Connection manager
Select from the list of existing Analysis Services connections, or create a new Analysis Services connection by using the New button described as follows.

New
Create a new connection by using the Add Analysis Services Connection Manager dialog box.

Mining structure
Select from the list of available mining structures, or create a new structure by clicking New.

New
Create a new mining structure and mining model by using the Data Mining Wizard.

Mining models
View the list of mining models associated with the selected mining structure.

Data Mining Model Training Editor (Columns Tab)

Use the Columns page of the Data Mining Model Training Editor dialog box to map input columns to columns in the mining structure.

Options

Available Input Columns
View the list of available input columns. Drag input columns to map them to mining structure columns.

Mining Structure Columns
View the list of mining structure columns. Drag mining structure columns to map them to available input columns.

Input Column
View input columns selected from the table above. To change or remove a mapping selection, use the list of Available Input Columns.

Mining Structure Columns
View each available destination column, whether mapped or not.