Databricks Exam Databricks Machine Learning Associate Topic 2 Question 27 Discussion

Actual exam question for Databricks's Databricks Machine Learning Associate exam

Question #: 27
Topic #: 2

[All Databricks Machine Learning Associate Questions]

A team is developing guidelines on when to use various evaluation metrics for classification problems. The team needs to provide input on when to use the F1 score over accuracy.

Which of the following suggestions should the team include in their guidelines?

AThe F1 score should be utilized over accuracy when the number of actual positive cases is identical to the number of actual negative cases.

BThe F1 score should be utilized over accuracy when there are greater than two classes in the target variable.

CThe F1 score should be utilized over accuracy when there is significant imbalance between positive and negative classes and avoiding false negatives is a priority.

DThe F1 score should be utilized over accuracy when identifying true positives and true negatives are equally important to the business problem.

Show Suggested Answer

Suggested Answer: D

If the new solution requires that each of the three models computes a prediction for every record, the time efficiency during inference will be reduced. This is because the inference process now involves running multiple models instead of a single model, thereby increasing the overall computation time for each record.

In scenarios where inference must be done by multiple models for each record, the latency accumulates, making the process less time efficient compared to using a single model.

Model Ensemble Techniques

by Omer at Mar 23, 2025, 10:38 AM

Limited Time Offer

25%

Off

Get Premium Databricks Machine Learning Associate Questions as Interactive Web-Based Practice Test or PDF

Contribute your Thoughts:

Submit Cancel

Michell

1 months ago

Haha, Cyril's analogy is spot on! It's like using a speedometer to judge a Formula 1 driver's performance. The F1 score is the true measure of a model's prowess in these tricky situations.

upvoted 0 times

...

1 months ago

Hmm, I was thinking B, but I can see how C would be more relevant. After all, who wants a model that's great at identifying true negatives but can't spot the true positives? The F1 score keeps things balanced.

upvoted 0 times

Eric

20 days ago

Definitely, C is the way to go. It's all about finding that balance between true positives and true negatives.

upvoted 0 times

...

Lashunda

1 months ago

Yeah, C seems like the best choice here. Balancing the trade-off between precision and recall is crucial.

upvoted 0 times

...

Chandra

1 months ago

I agree, C makes more sense. It's important to prioritize avoiding false negatives in certain cases.

upvoted 0 times

...

2 months ago

Option C seems like the best choice here. When there's a significant imbalance between positive and negative classes and avoiding false negatives is important, the F1 score is more appropriate than just looking at overall accuracy.

upvoted 0 times