Exam DP-100 topic 1 question 37 discussion

I feel like like the two requirements are conflicting: every possible combination implies Entire grid while lower computational resources implies Random grid. "Entire grid: When you select this option, the component loops over a grid predefined by the system, to try different combinations and identify the best learner. This option is useful when you don't know what the best parameter settings might be and want to try all possible combinations of values." "Random sweep: When you select this option, the component will randomly select parameter values over a system-defined range. You must specify the maximum number of runs that you want the component to execute. This option is useful when you want to increase model performance by using the metrics of your choice but still conserve computing resources."

All combination + Reduce computing resource , because "Research has shown that this method (Random Grid Sweep) yields the same results, but is more efficient computationally." I think D would be the best choice

_ALLOWS_ for every possible combination doesn't mean it has to iterate through all. The requirement to reduce compute resources obviously points us away from a full grid search.

Vote for Entire Grid Sweep Mode since one requirement is every possible combination of hyperparameters to be iterated. Other options reduce computational resources yet do not satisfy this requirement.

"Maximum number of runs on random grid: This option also controls the number of iterations over a random sampling of parameter values, but the values are not generated randomly from the specified range; instead, a matrix is created of all possible combinations of parameter values and a random sampling is taken over the matrix. This method is more efficient and less prone to regional oversampling or undersampling."

The answer is D. Random seed ALLOWS every possible combination. It won't go through every possible combination but any random combination is possible.

Question says "You have to make sure that the sweep allows for every possible combination of hyperparameters to be iterated". There is no way to guarantee Random sweep will get all possible combinations, answer is C

D. You should consider making use of the Random grid sweep mode. The Random grid sweep mode randomly selects combinations of hyperparameters to test, reducing the number of total combinations and the computing resources needed to carry out the sweep. This method can still provide a good understanding of the relationship between hyperparameters and model performance, but may require multiple runs to converge on the optimal hyperparameters.

If we want every possible combi9nation we need entire grid, so C is correct.

C is correct..

Keyword -> "sweep allows for every possible combination" so, option B https://learn.microsoft.com/en-us/azure/machine-learning/component-reference/tune-model-hyperparameters

Entire grid is the only one can try all combinations, but random sweep is low computational cost

Related with this question https://www.examtopics.com/discussions/microsoft/view/43145-exam-dp-100-topic-2-question-80-discussion/

Answer D. there are two reqs... entire grid and low computational cost

Random grid

ans: C

"every possible combination", hence Entire grid sweep mode