You had the Machine Learning activated on your system. However, when you create a model and try to activate it, you get the message "Model activation failed".
SAP Cloud for Customer
Reproducing the Issue
Pre-requirement: You first need to open an incident requesting to have the Machine Learning activated in your system.
- Go to the Administration work center.
- Go to the Prediction Services view.
- Click Model Setup.
- Select your Machine Learning scenario.
- Click Add Model.
- Click Activate.
RESULT: You get a message "Model activation failed".
This message can be triggered for different reasons. Two options are:
- Only one model can be active at any time. If you already have another model active, you need to deactivate one to activate the other.
- If you only have one model active: The criteria for activating a model is that the "Predictive Power" is more than 50% AND the "Prediction Confidence" more than 95%. You can check these parameters by clicking "Performance".
The most probable cause for your Predictive Power being lower than 50% is the data quality. You can get an idea about the data quality while training the model. You can also get a glimpse of the data quality by looking at the Readiness Report.
It is not possible to manually change/increase either of these two features ("Predictive Power" and "Prediction Confidence").
- Administrator Guide - Machine Learning
- 2643651 - Readiness Check Not Available For Machine Learning Scenarios
- 2772877 - Machine Learning: Readiness check failure
- 2659267 - Some Opportunities are not Scored by Machine Learning
Machine learning, report, readiness, train, model, Model activation failed , KBA , LOD-CRM-ML , C4C Machine Learning , LOD-CRM-OPP , Opportunity Management , How To