Skip to main content


         This documentation site is for previous versions. Visit our new documentation site for current releases.      
 

Supported models for import

Updated on July 5, 2022

Learn more about the PMML and H2O models that you can import to Prediction Studio.

Table of contents

This article covers the following topics:

  1. Supported PMML model types
  2. Guidelines and restrictions for importing PMML models
  3. Unsupported PMML model types
  4. Supported H2O model types

Supported PMML model types

Pega Platform uses a specific implementation of the PMML format, which means that some PMML features and models are not supported in the Predictive Model rule.

You can import models from the following PMML versions:

  • 3.0
  • 3.1
  • 3.2
  • 4.0
  • 4.1
  • 4.2
  • 4.2.1
  • 4.3
  • 4.4

You can import PMML models that use the following algorithms:

  • Anomaly detection
  • Clustering
  • Decision tree
  • General regression
  • K-nearest neighbors
  • Naive Bayes
  • Neural network
  • Regression
  • Ruleset
  • Scorecard
  • Support Vector Machine
  • Ensemble methods (including Random forest and Gradient boosting)

Guidelines and restrictions for importing PMML models

When importing PMML models to Pega Platform, take into account the following guidelines and restrictions:

  • Clustering:
    • The kind attribute of the ComparisonMeasure element can be set to distance or similarity.
  • Decision tree:
    • The functionName attribute for the TreeModel element cannot be empty, and has to be set to regression or classification.
  • General regression:
    • If the functionName attribute of the GeneralRegressionModel element is regression, the model must have exactly one PPMatrix.
    • The multinomialLogistic, ordinalMultinomial, and CoxRegression algorithms are not supported for the regression mining function.
    • The regression, general_linear, and CoxRegression algorithms are not supported for the classification mining function.
  • K-nearest neighbors:
    • The opType attribute of the input field (DataField) can be set to continuous or categorical.
    • The kind attribute of the ComparisonMeasure element can be set to distance or similarity.
  • Naive Bayes models support only one classification mining function.
  • Neural network:
    • The mining function can be regression or classficiation.
  • Regression:
    • If the functionName attribute of the RegressionModel element has the value regression, the normalizationMethod attribute can have one of the following values: none, softmax, logit, or exp.
    • If the functionName attribute of the RegressionModel element has the value classification, the normalizationMethod attribute can have one of the following values: none, softmax, logit, loglog, or cloglog.
  • Scorecard:
    • The functionName attribute is mandatory for the ScorecardModel element.
  • Support Vector Machine:
    • The svmRepresentation attribute is mandatory for the SupportVectorMachineModel element.
    • The functionName attribute for the SupportVectorMachineModel element cannot be empty and has to be set to regression or classification.
    • The probability attribute value is not supported for the resultFeature attribute in the Output element.

Unsupported PMML model types

Pega Platform does not support PMML models that use the following algorithms:

  • Association rules
  • Base line
  • Ensemble methods (including Mining and Many-in-one) that contain composite embedded models
  • Sequences
  • Text
  • Time series

Supported H2O model types

You can import H2O-3 and Driverless AI models in .mojo format.

You can import H2O-3 models that use the following algorithms:

  • Cox proportional hazards
  • Deep learning
  • Distributed random Forest
  • Generalized linear model
  • Gradient boosting machine
  • Isolation forest
  • K-means clustering
  • Naive Bayes classifier
  • Stacked ensembles
  • Support Vector Machine
  • XGBoost

Have a question? Get answers now.

Visit the Support Center to ask questions, engage in discussions, share ideas, and help others.

Did you find this content helpful?

Want to help us improve this content?

We'd prefer it if you saw us at our best.

Pega.com is not optimized for Internet Explorer. For the optimal experience, please use:

Close Deprecation Notice
Contact us