nerdexam
Amazon

MLS-C01 · Question #389

A data scientist uses Amazon SageMaker to perform hyperparameter tuning for a prototype machine leaming (ML) model. The data scientist's domain knowledge suggests that the hyperparameter is highly…

The correct answer is D. Reverse logarithmic scaling. Reverse logarithmic scaling is the appropriate choice when the optimal value of a hyperparameter is expected to be closer to the upper end of the range, in this case, closer to 1.0. Reverse logarithmic scaling gives more granularity to values near the upper bound of the…

Modeling

Question

A data scientist uses Amazon SageMaker to perform hyperparameter tuning for a prototype machine leaming (ML) model. The data scientist's domain knowledge suggests that the hyperparameter is highly sensitive to changes. The optimal value, x, is in the 0.5 < x < 1.0 range. The data scientist's domain knowledge suggests that the optimal value is close to 1.0. The data scientist needs to find the optimal hyperparameter value with a minimum number of runs and with a high degree of consistent tuning conditions. Which hyperparameter scaling type should the data scientist use to meet these requirements?

Options

  • AAuto scaling
  • BLinear scaling
  • CLogarithmic scaling
  • DReverse logarithmic scaling

How the community answered

(13 responses)
  • A
    8% (1)
  • B
    8% (1)
  • D
    85% (11)

Explanation

Reverse logarithmic scaling is the appropriate choice when the optimal value of a hyperparameter is expected to be closer to the upper end of the range, in this case, closer to 1.0. Reverse logarithmic scaling gives more granularity to values near the upper bound of the specified range (0.5 < x < 1.0), allowing the model to more efficiently explore this area with fewer tuning runs. This is particularly useful when domain knowledge suggests sensitivity in that specific region. Auto scaling automatically determines scaling but may not focus on the upper end of the range, which is crucial here. Linear scaling treats all values in the range equally, which might not efficiently target the area close to 1.0. Logarithmic scaling gives more focus to values near the lower end of the range, which is the opposite of what is needed in this case.

Topics

#Hyperparameter Tuning#Amazon SageMaker#Scaling Types#Model Optimization

Community Discussion

No community discussion yet for this question.

Full MLS-C01 Practice