Using PMML Models with InterSystems Products

You can execute PMML models in InterSystems IRIS by loading the model file and interacting with it via the Management Portal interface or a programmatic API to run predictions and scoring.

InterSystems IRIS supports PMML versions 4.1 through 4.4.1, enabling execution of predictive models including Naive Bayes, Regression, Tree, SVM, Clustering, RuleSet, and Neural Networks.

Yes, integrating PMML models into InterSystems IRIS bridges data science and production environments, allowing you to deploy trained models directly into production workflows for predictive analytics.

Yes, InterSystems IRIS provides a programmatic API for PMML model interaction, enabling you to automate predictive scoring and integrate predictions seamlessly within your applications.

The best way to test a PMML model against sample data is using the InterSystems IRIS PMML Model Tester, which allows you to quickly load and execute models to verify scoring outputs.

Using PMML for predictive analytics in InterSystems IRIS eliminates the gap between model training and deployment, allowing direct in-database execution of data science models without external scoring services.