Answer-first summary for fast verification
Answer: You should consider making use of the Random grid sweep mode.
The question presents conflicting requirements: ensuring every possible combination of hyperparameters can be iterated while reducing computational resources. The community discussion reveals that 'Entire grid sweep' (C) guarantees all combinations but is computationally expensive, while 'Random grid sweep' (D) reduces resources but doesn't guarantee all combinations. However, the consensus (68% for D, with high upvotes on comments supporting D) favors D because the requirement states the sweep must 'allow for' every possible combination, not that it must iterate through all. Random grid sweep can theoretically sample any combination from the full parameter space while being computationally efficient, making it the optimal compromise. Comments note that research shows random sweep yields similar results with better efficiency, and it's specifically designed to conserve computing resources while exploring the parameter space.
Author: LeetQuiz Editorial Team
Ultimate access to all questions.
You are creating a binary classification model in Azure Machine Learning Studio. You need to perform a hyperparameter tuning sweep that iterates over every possible combination of parameters while minimizing the computational resources required.
Which parameter sampling method should you use?
A
You should consider making use of the Selective grid sweep mode.
B
You should consider making use of the Measured grid sweep mode.
C
You should consider making use of the Entire grid sweep mode.
D
You should consider making use of the Random grid sweep mode.
No comments yet.