How To Fix RSDME236 - Model field &1 has untrained value

SAP Error Message - Details

• Message type: E = Error

• Message class: RSDME -

• Message number: 236

• Message text: Model field &1 has untrained value

What causes this issue?

The specified model field has values that was not found in the training
data or not considered for a split in the decision tree.

System Response

If this model field is considered for a split in the decision tree,
then only those values on which the split has occured would be
considered valid. Any other value will be treated as missing/unknown.

How to fix this error?

If the specified value was intentional, ignore this warning.
If you get this message in the online prediction screen, then choose
the possible entries function(F4 help) for that model field.

Procedure for System Administrators

The system issues an error message and will not allow you to continue with this transaction until the error is resolved.

Error message extract from SAP system. Copyright SAP SE.

Smart SAP Assistant

• What is the cause and solution for SAP error message RSDME236 - Model field &1 has untrained value ?

  The SAP error message RSDME236, which states "Model field &1 has untrained value," typically occurs in the context of SAP BW (Business Warehouse) or SAP HANA when working with predictive analytics or machine learning models. This error indicates that the model is encountering a value in the specified field that it has not been trained on, which can lead to issues in data processing or analysis.
  Cause:
  
  Untrained Values: The model was trained on a specific set of values, and the incoming data contains values that were not present during the training phase.
  Data Quality Issues: There may be inconsistencies or errors in the data being fed into the model, such as typos, missing values, or unexpected categories.
  Model Configuration: The model may not be configured correctly to handle all possible values for the specified field.
  Changes in Data: If the underlying data has changed (e.g., new categories added), the model may need to be retrained to accommodate these changes.
  
  Solution:
  
  Data Validation: Check the incoming data for the field specified in the error message. Ensure that all values are valid and consistent with what the model was trained on.
  Retrain the Model: If new values are present in the data that were not included during the initial training, consider retraining the model with the updated dataset that includes these new values.
  Handle Untrained Values: Implement a strategy to handle untrained values, such as:
  
  Mapping untrained values to a default category.
  Excluding records with untrained values from the analysis.
  Using imputation techniques to replace untrained values with a suitable alternative.
  
  
  Model Configuration Review: Review the model configuration to ensure it is set up to handle the expected range of values for the specified field.
  Documentation and Logging: Maintain documentation of the model's training data and log any occurrences of untrained values to monitor and address them in future iterations.
  
  Related Information:
  
  Model Training: Understand the process of training machine learning models in SAP BW or HANA, including data preparation, feature selection, and validation.
  Data Governance: Implement data governance practices to ensure data quality and consistency across the organization.
  SAP Notes and Documentation: Check SAP Notes and official documentation for any updates or patches related to this error message, as well as best practices for model training and deployment.
  Community Forums: Engage with SAP community forums or support channels for additional insights and solutions from other users who may have encountered similar issues.
  
  By addressing the root cause of the untrained values and ensuring that the model is properly configured and trained, you can resolve the RSDME236 error and improve the reliability of your predictive analytics processes in SAP.