Creating a Training model and Building Data Mining Model Using OLAP Data store.

In the section here we will build a Data Mining model using OLAP data store.

1. In the Analysis Manager Tree pane expand the FoodMart 2000 database.
2. Right click on Mining models and select New Mining model… option.

3. The Data mining Wizard opens. Click Next on the Welcome screen to proceed with the creation of the data mining model

4. The next screen prompts the user select the Data source type. Select OLAP and click Next.

5. Now select the Algorithm for this model. Since we want to use Clustering Algorithm select it and click Next to continue.

6. Since we are studying customer behavior we will chose the sales cube for our analyses. Click on Sales and click Next

7. By default the lowest level would be store and the level would be Name. We will select Customers as our dimension and Name as the level. Click Next after selecting the above.

8. The Next step is the selection of the Training data. The Wizard defaults to the selection of the training data from the dimension which was selected in the previous screen. Click Next to continue

9. Now give the model a name “Customer Sales” and Save and process the model by selecting the appropriate radio button. Click finish to process the mining model.

10. The Process log window opens and the Mining model is processed. On successful completion of processing a message is displayed. Close the Process log window. It may be pointed out here that the process log window displays the location where the mining algorithm is being processed. The process can be stopped any time and rebuilt. The user can reprocess the Mining model or simply exit without processing at any time.

Next Page: Tutorial 67: MSAS - Introduction to Data Mining - Page 3