Besides providing a new predicted value, which additional task would some modeling nuggets perform?

Question

Options

• AProvide additional information regarding the quality of the prediction.
• BFilter non-target fields from being passed downstream.
• CUse the scored data to update itself to increase its accuracy.
• DExport the predictions to external storage or processes for further use.

• A
  3% (1)
• B
  3% (1)
• C
  89% (31)
• D
  6% (2)

Explanation

Some modeling nuggets - particularly those supporting online or adaptive learning (such as SPSS Modeler's Self-Learning Response Model) - can feed scored data back into themselves to incrementally retrain and improve over time. This self-updating behavior is what makes C correct: beyond generating a prediction, these nuggets treat each scored record as new training input, closing the feedback loop automatically.

Why the distractors are wrong:

• A is tempting but describes something nearly all modeling nuggets do (e.g., confidence scores) - not a capability unique to "some."
• B describes field filtering, which is handled by other node types (like a Filter node), not a modeling nugget's scoring function.
• D describes output/export behavior, which is the job of Export nodes downstream, not the modeling nugget itself.

Memory tip: Think of the word "some" in the question as your signal. Most nuggets score; only some also learn. If a nugget can update itself, it's doing more than predicting - it's adapting. Associate "self-updating model = online learning nugget" and you'll lock in C every time.

No community discussion yet for this question.