Predictive modeling tutorials

The tutorials in this section provide detailed examples of how you can build predictive models and integrate them into your decisions.
Some of the tutorials provide instructions for using samples that are available in Decision Intelligence:
Tutorial Provides instructions to use
Using machine learning to make better decisions Machine learning complete tutorial - Loan approval
Quick guide: Integrating Watson Machine Learning with decision services Machine learning short tutorial - Loan approval
Quick guide: Using the embedded machine learning provider Machine learning quick tutorial - Loan approval
Using machine learning to reduce customer churn Machine learning sample - Customer loyalty

These samples can be imported from the New decision service wizard. For more information about importing samples, see Building decision services.

Setting up Industry samples

Decision Intelligence also includes two industry-specific samples: Banking and Telecom. When importing these samples, a warning appears. You can acknowledge the warning and manually configure the predictive models by following the instructions in Quick guide: Using the embedded machine learning provider.
Sample PMML file
Banking Use the ML-Sample-Classifier-StandardScaler-pmml.xml file to configure the Loan risk score model available in the Approval with ML decision service.
Telecom
  • Use the Churn_RandomForestClassifier.xml file to configure the Customer Churn model available in the Retention with ML decision service.
  • Use the LTV_Regression.xml file to configure the Customer Lifetime Value model available in the Retention with ML decision service.