Google Professional Machine Learning Engineer Exam - Topic 4 Question 24 Discussion

Actual exam question for Google's Professional Machine Learning Engineer exam

Question #: 24
Topic #: 4

[All Professional Machine Learning Engineer Questions]

You work for a social media company. You need to detect whether posted images contain cars. Each training example is a member of exactly one class. You have trained an object detection neural network and deployed the model version to Al Platform Prediction for evaluation. Before deployment, you created an evaluation job and attached it to the Al Platform Prediction model version. You notice that the precision is lower than your business requirements allow. How should you adjust the model's final layer softmax threshold to increase precision?

AIncrease the recall

BDecrease the recall.

CIncrease the number of false positives

DDecrease the number of false negatives

Show Suggested Answer

Suggested Answer: D

by Nickolas at May 06, 2022, 03:52 AM

Limited Time Offer

25%

Off

Get Premium Professional Machine Learning Engineer Questions as Interactive Web-Based Practice Test or PDF

Contribute your Thoughts:

Submit Cancel

Cherry

4 months ago

Just adjusting the threshold won't fix everything, right?

upvoted 0 times

...

Kaycee

4 months ago

I thought increasing false positives was a good strategy?

upvoted 0 times

...

Celia

4 months ago

Wait, isn't increasing recall the opposite of what we want?

upvoted 0 times

...

Audry

4 months ago

Totally agree, D is the way to go!

upvoted 0 times

...

Darci

5 months ago

You need to decrease false negatives to boost precision.

upvoted 0 times

...

Mari

5 months ago

I vaguely remember that decreasing false positives is usually linked to increasing precision, but I'm not confident about how that relates to the softmax threshold adjustment.

upvoted 0 times

...

Freeman

5 months ago

I feel like increasing recall could be a trap here. I need to double-check if that actually helps with precision in this context.

upvoted 0 times

...

Graciela

5 months ago

I think we practiced a similar question where we had to balance precision and recall. If I recall correctly, decreasing false negatives might help improve precision.

upvoted 0 times

...

Gerry

5 months ago

I remember we discussed how adjusting the softmax threshold can impact precision and recall, but I'm not sure which direction to go for increasing precision.

upvoted 0 times

...

Meghan

5 months ago

Based on my understanding of the different AEM as a Cloud Service roles, I believe the DevOps role is the one responsible for managing and running pipelines. I'll select option B.

upvoted 0 times

...

Devora

5 months ago

I'm a little confused by this question. The options don't seem to be directly related to improving the use of tools and techniques. I'll have to think about this one a bit more.

upvoted 0 times

...

Hyun

5 months ago

Hmm, I'm a bit unsure about this one. I'll need to carefully read through each option and think about how they differ in terms of defining a test case.

upvoted 0 times

...

Pearline

5 months ago

Hmm, the question mentions the load balancing method needs to account for the pool member's response over time. That sounds like it might rule out the Fastest option.

upvoted 0 times

...

Rozella

5 months ago

Wait, I'm confused. Is it about the organization being ready to accept the solution, or the performance evaluation being positive? I need to re-read this carefully.

upvoted 0 times

...