Load the Netflix customer churn dataset using the CSV Reader
The Missing Value node imputes any incomplete records. The Number to String node converts the target variable 'churned' from an integer to a nominal string for classification. The Domain Calculator ensures all categorical features have a valid domain. (Note: The Column Filter removed 'customer_id' to prevent model overfitting
The dataset is split into two disjoint sets: 80% for training the model and 20% for testing.
A Random Forest Learner is trained on the 80% training partition to predict 'churned'. The Predictor applies this model to the 20% test set. Finally, the Scorer generates the Confusion Matrix, Accuracy, Sensitivity, and Specificity metrics to evaluate model performance.
To use this workflow in KNIME, download it from the below URL and open it in KNIME:
Deploy, schedule, execute, and monitor your KNIME workflows locally, in the cloud or on-premises – with our brand new NodePit Runner.
