Oracle® Fusion
Applications Workforce Deployment Implementation Guide 11g Release 6 (11.1.6) Part Number E20379-06 |
![]() Home |
![]() Contents |
![]() Book List |
![]() Contact Us |
![]() Previous |
![]() Next |
This chapter contains the following:
Managing Predictive Models: Explained
Oracle Fusion Workforce Predictions provides predefined models for the prediction of worker performance and voluntary termination. Each predictive model is based on multiple attributes.
You can:
Run predictive models to provide up-to-date predictions.
Remove individual predictive models from the predictions process.
Remove individual attributes from the predictive models or what-if analyses.
Create predictive attributes to include in the predefined predictive models or what-if analyses.
When you run a predictive model, the process Collect Data and Perform Data Mining for Predictive Analytics is invoked immediately to:
Rebuild the selected predictive models.
Make predictions based on scores derived during the build process.
If the volume of relevant transactions (such as transfers, hires, terminations, and promotions) is high in your enterprise, then you need to schedule the process Collect Data and Perform Data Mining for Predictive Analytics to run weekly. At a minimum, you are recommended to run the process monthly to take account of latest data trends. When scheduled, the process rebuilds and runs all predictive models.
If you add attributes to or remove attributes from a predictive model, and you want to include those changes in predictions immediately, then you need to run the predictive model immediately rather than wait for the next scheduled run of Collect Data and Perform Data Mining for Predictive Analytics.
To remove a predictive model from the predictions process, you deselect the Include in Predictions option for the model. In this case, the model is excluded when you run Collect Data and Perform Data Mining for Predictive Analytics, whether you run it immediately or as a scheduled process. Consequently, related analytics in transactional flows, such as Promote Worker, are empty.
You can create predictive attributes to include in the predefined predictive models. To derive the value of the new attribute, you create a fast formula database item (DBI) group and select it in the Formula Function field. You can also control which predefined and locally created predictive attributes appear in what-if analyses.
In Saas (Software as a Service) environments, you cannot create formula functions; therefore, you may not be able to create predictive attributes.
You can edit or delete any predictive attribute that you create; you cannot edit or delete predefined predictive attributes. For any attribute, you can edit how the attribute appears in what-if analyses. For example, you can change the minimum and maximum values on a slider scale.