Final Workflow

Churn Prediction - Solution to the exercise use case for KNIME User Training
- Read data from different sources
- Concatenate and join data tables
- Bring values from a reference table to the data
- Modify String values, change column types, and handle missing values
- Visualize data to explore it
- Train a decision tree model, and apply it to test data and real data
- Evaluate the preformance of the classification model
- Export the decision tree model as a PMML model file
- Write the prediction results into a database
- Generate a BIRT report that shows the results

External resources

• Slides (KNIME Analytics Platform Course)
• Drag&Drop Data Science