Icon

Building a Churn Prediction Model

Training a Churn Predictor

This workflow is an example of how to build a basic PMML model for a churn prediction using a Decision Tree algorithm.

Churn Prediction: TrainingThis workflow illustrates how to build and evaluate a churn prediction model, i.e. predict the likelihood that a customer will churn a specific contract.Task Train and evaluate a model to predict customer churn. The workflow is also available in the EXAMPLES server under 50_Applications/_18_Churn_PredictionUse Case is described at URL: https://www.knime.org/knime-applications/churn-predictionImplementation is described in this KNIMETV video "Building a basic Model for Churn Prediction with KNIME" https://youtu.be/RHsO10q7e2Y Data Reading Two files: - contract data + churn - behavioral (calls) dataBoth files are located inTheData/Customers Pre-processing - Join contract data and behavioral data - Convert Churn values to String to be used as class in upcoming classification - Color rows by churn value - Reserve 80% of the rows for model training and remaining for model testing - Use same number of data rows for both classes in testing Score the ModelRemember to use the Predictornode appropriate for your model!Evaluate predictions based onconfusion matrix and ROC. Train a ModelThe Decision Tree here is just an example. Open the Decision Tree Learner view to see thedecision tree paths. training set test set Churn = 0 customerremained with contractChurn = 1 customer quitcontract 80% vs. 20%Apply the trained modelto predict ChurnMatch original vs. predicted Churn valuesColor rows by churn valueChurn = 0 -> BlueChurn = 1 -> RedSave model in PMML formatReadingContractData.csvJoin the contract data and the behavioral dataTarget Class = ChurnBuild ROC for predictionsReading CallsData.xlsArea codeand churn 0/1are converted toString. Partitioning Decision TreePredictor Scorer Color Manager PMML Writer File Reader Joiner DecisionTree Learner ROC Curve(JavaScript) Excel Reader (XLS) Number To String Churn Prediction: TrainingThis workflow illustrates how to build and evaluate a churn prediction model, i.e. predict the likelihood that a customer will churn a specific contract.Task Train and evaluate a model to predict customer churn. The workflow is also available in the EXAMPLES server under 50_Applications/_18_Churn_PredictionUse Case is described at URL: https://www.knime.org/knime-applications/churn-predictionImplementation is described in this KNIMETV video "Building a basic Model for Churn Prediction with KNIME" https://youtu.be/RHsO10q7e2Y Data Reading Two files: - contract data + churn - behavioral (calls) dataBoth files are located inTheData/Customers Pre-processing - Join contract data and behavioral data - Convert Churn values to String to be used as class in upcoming classification - Color rows by churn value - Reserve 80% of the rows for model training and remaining for model testing - Use same number of data rows for both classes in testing Score the ModelRemember to use the Predictornode appropriate for your model!Evaluate predictions based onconfusion matrix and ROC. Train a ModelThe Decision Tree here is just an example. Open the Decision Tree Learner view to see thedecision tree paths. training set test set Churn = 0 customerremained with contractChurn = 1 customer quitcontract 80% vs. 20%Apply the trained modelto predict ChurnMatch original vs. predicted Churn valuesColor rows by churn valueChurn = 0 -> BlueChurn = 1 -> RedSave model in PMML formatReadingContractData.csvJoin the contract data and the behavioral dataTarget Class = ChurnBuild ROC for predictionsReading CallsData.xlsArea codeand churn 0/1are converted toString. Partitioning Decision TreePredictor Scorer Color Manager PMML Writer File Reader Joiner DecisionTree Learner ROC Curve(JavaScript) Excel Reader (XLS) Number To String

Nodes

Extensions

Links