You use the Azure Machine Learning SDK to run a training experiment that trains a classification model and calculates its accuracy metric. The model will be retrained each month as new data is availab

Question

Options

• ASpecify a different name for the model each time you register it.
• BRegister the model with the same name each time regardless of accuracy, and always use the
• CSpecify the model framework version when registering the model, and only register subsequent
• DSpecify a property named accuracy with the accuracy metric as a value when registering the
• ESpecify a tag named accuracy with the accuracy metric as a value when registering the model,

• A
  4% (1)
• B
  8% (2)
• C
  83% (20)
• D
  4% (1)

Explanation

Two approaches allow conditional registration based on accuracy: (E) Register the model with a tag named 'accuracy' set to the metric value. Before registering a new model, retrieve the currently registered model's accuracy tag, compare it to the new model's accuracy, and only call register() if the new value is higher. Tags are mutable key-value pairs that are queryable via the model registry. (C) A similar approach using model properties or framework versioning to track accuracy-properties are immutable once set at registration, providing a reliable record. Option D (property named accuracy) is logically equivalent to E but uses immutable properties; the exam key credits C and E. Options A and B do not provide a mechanism to conditionally register based on accuracy comparison.

No community discussion yet for this question.