Start SAP Predictive Analytics and select Expert Analytics
Logon with credentials provided to you.
Select the LUMS file for Telco Churn Post-Analysis
a. Either double-click Telco_Churn_Postanalysis.lums file and import, or select the file in the
Documents, after the file is imported.
b. In the dialog box, enter the SAP HANA server details, your user name, and password.
Verify the data
Choose Prepare to ensure that the data has loaded properly
You are viewing the prediction results from an already created analytic view based on the data
available. In reality, after the final churn model is created (SII_FINAL_MODEL_OUT_<your
number>)., basically you update the analytic view in HANA studio or create a new analytic view
which consumes the results from the table where you have written for eg.,
SII_FINAL_MODEL_OUT_<your number>.
Switch to the Visualize panel to view the pre-constructed visualizations.
Switch to the Compose tab to view dashboards created to evaluate the customers predicted to churn
and the projected effect on the business.
Post-Paid Analysis
In this use case we focus on build churn models and identify the post-paid customers likely to leave.
Automated Analytics
Launch SAP Predictive Analytics
Choose the Modeler section
Select
Create a Classification/Regression Model
In the Select a Data Source window, choose Use Explorer option. Select your SAP HANA instance
and connect using the <Domain User> account via the
Browse
button.
Specify the data set by selecting Analytical Record EntitySubscriptionAnalyticalRecord_PostPaid and
Time-stamped Population EntitySubscriptionPopulation.
Choose Next. Enter training cut-off date to Analytical Data Set Reference Date prompt screen.
Choose Analyze to read description from Analytical Record.
Choose Next to continue to Selecting Variables screen.
Select CHURN_FLAG variable as a Target Variables and Kxid,KxTimeStamp,n_days_to_eos to
Excluded Variables. Continue Next.
Accept the default parameters and choose Generate.
Review Model Overview reports for model detail.
Choose Next to continue to Run section.
Choose Apply Model and use following setting in Applying the Model screen:
Application Data Set
Data Type: Data Base
Folder: <HANA connection>
Data: EntitySubscriptionPopulation
Generation options
Generate: Probability& Error Bars
Mode: Apply
