Introduction to Java Model Scoring

Java Model Scoring enables you to score new data using the formula created when you fit a machine learning model.

Note: Models must first be exposed through Machine Learning Model Management before they become available in the Java Model Scoring stage. For more information see Introduction to Model Assessment.

To score your data, you must complete two tabs of the Java Model Scoring Options dialog. First identify the model and its type, then ensure the model's fields are correctly mapped to Spectrum™ Technology Platform fields. Following that, you configure the output by selecting which fields you want to include and running your job. The Model Output tab contains mapping for data types for Spectrum™ Technology Platform and your model.

If your job contains a stage that captures the output in a file or a table, you can use that ouptut in a subsequent dataflow or web service.