Creating a predictive model

Data scientists use customer data to develop powerful and reliable models that can predict customer behavior, such as offer acceptance, churn rate, credit risk, or other types of behavior.

When creating predictive models in Pega Platform, data scientist can create new predictive models in the Decision Analytics work area to be used only in Pega Platform or import third-party models in Predictive Model Markup Language (PMML) format. PMML is an open standard that facilitates sharing and reusing predictive models that were not created in Pega Platform.

1. Open the Decision Analytics portal from Designer Studio.

2. Click Create > Predictive model.

3. Enter a name for your model and select one of the following options.

   Build new model

   1. Select a category and a model template.

      Note:  When model templates are not available, you need to import them in the Decision Analytics Settings landing page.

   2. Click Start.

   3. Optional: Edit settings.

   4. Continue with Preparing data.

   Import PMML

   1. Click Choose and select a model file to upload.
      The PMML supported models

      • Clustering
      • GeneralRegressionModel
      • MiningModel
      • NaiveBayesModel
      • NearestNeighborModel
      • NeuralNetwork
      • RegressionModel
      • RuleSetModel
      • Scorecard
      • SupportVectorMachineModel
      • TreeModel
   2. Use the default context or specify a context.
   3. Click Ruleset and specify details.
   4. Click Branch and specify details.
   5. Click Import.
   6. Continue model configuration in the Predictive Model form. For more information, see About Predictive Model rules.

Open topic with navigation