Skip to main content


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

Improve prediction accuracy by monitoring predictive models (8.2)

Updated on May 3, 2021

You can improve predictions of your customer needs and decisions by closely monitoring the predictive models that you use, for example, in customer-oriented strategies.

Whether you build your own model or import an existing one, you can analyze and compare both types by using advanced statistical metrics. If you decide to create a model, you do not need to define the outcomes to monitor because the Pega Platform™ templates already contain their standard definitions; if you import a preconfigured model from a PMML file, you can specify the outcome that you want to monitor by using an improved import feature:

Thumbnail
Sample PMML model outcome definition options

After defining the outcome that you want to monitor and running a model to gather responses over time, you can use detailed metrics to verify the model performance:

Monitoring a predictive model

Additionally, you can compare the performance of all predictive models of the same type in summarized reports:

Viewing predictive model reports

For more information, see:

  • Previous topic Create custom criteria for Proposition Filter rules by using the condition builder (8.3)
  • Next topic Interpret the decision funnel with simulation tests (8.2)

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