Error while training solution, details: NSE0016:Failed to execute component Preprocessor : NCE0079:Input data is null or emptyIssue While training "Solution Definitions", sometimes they will fail with the message "NSE0016:Failed to execute component Preprocessor : NCE0079:Input data is null or empty". Steps to train a Solution: Open a list of Solution Definitions using the left hand navigation menuOpen one of the Solution Definition recordsClick "Update & Retrain" in the header of the formA new record will appear in the "ML Solutions" related list of that Solution DefinitionThe "State" field of the ML Solution will show the progress If your Solution fails due to an error, the "State" field will say "Error while training solution". Steps to see the error details: Open the failed record from the "ML Solutions" listClick "Show training progress" You may see a message like this: ReleaseAll releases.CauseThe message indicates there are two failures. The first message, "NSE0016:Failed to execute component Preprocessor" means that there was a failure during Preprocessing. This is the phase of ML Training where automated checks are executed against the Data Set to ensure the data is valid. If invalid data is found in a field, the entire column is omitted from training. There are many possible causes of Preprocessor failures including the following: Some of the data is missing/blank.The data is highly unique - data must not be more than 99.3% unique.In older versions of Predictive Intelligence the value 0 was considered invalid in some cases. For example, the ML Solution.Solution Type "Classification" did not allow the integer value 0 or string value "0" to be used in Training. That Solution Type has been deprecated but some customers may have ML Solutions that were trained with the older type. The second message, "NCE0079:Input data is null or empty" means that, after columns have been omitted from the training, there are no columns left to train the model. Note that if a column is excluded from a Training Set due to having invalid values, your Training may still complete successfully if there is at least one other column in the Training Set that does not contain any invalid values.ResolutionCheck your data to see if any of the fields have empty values. Exclude or fill any empty rows. Check your data to see if it is highly unique. It must not be more than 99.3% unique. If running a the old non-workflow Classification solution type (this is a deprecated solution type), you must not have any string or integer fields with the value 0.