Google Professional Machine Learning Engineer Exam - Topic 7 Question 29 Discussion

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

Question #: 29
Topic #: 7

[All Professional Machine Learning Engineer Questions]

You were asked to investigate failures of a production line component based on sensor readings. After receiving the dataset, you discover that less than 1% of the readings are positive examples representing failure incidents. You have tried to train several classification models, but none of them converge. How should you resolve the class imbalance problem?

AUse the class distribution to generate 10% positive examples

BUse a convolutional neural network with max pooling and softmax activation

CDownsample the data with upweighting to create a sample with 10% positive examples

DRemove negative examples until the numbers of positive and negative examples are equal

Show Suggested Answer

Suggested Answer: B

by Jolene at May 03, 2022, 07:09 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

Carla

4 months ago

Convolutional networks might not be the best fit here.

upvoted 0 times

...

Britt

4 months ago

Wait, removing negatives? That seems extreme!

upvoted 0 times

...

Janey

4 months ago

Generating more positive examples sounds like a solid plan.

upvoted 0 times

...

Justine

4 months ago

I think downsampling is risky, might lose important data.

upvoted 0 times

...

Hershel

5 months ago

Class imbalance is a huge issue in ML!

upvoted 0 times

...

Glen

5 months ago

Using a CNN sounds interesting, but I don't recall it being a common solution for class imbalance issues.

upvoted 0 times

...

Sherron

5 months ago

I think downsampling could work, but I'm uncertain about how effective it would be in this case.

upvoted 0 times

...

Elmira

5 months ago

I remember reading about class imbalance, but I'm not sure if generating more positive examples is the best approach.

upvoted 0 times

...

Ethan

5 months ago

I practiced a similar question where we had to balance classes, and I think removing negatives might lead to losing important data.

upvoted 0 times

...

Carissa

5 months ago

I'm feeling pretty confident about this one. The drone control should run in the edge cloud for low latency, and the video app can run in the central cloud since it's less sensitive to latency.

upvoted 0 times

...

Felicidad

5 months ago

This looks like a tricky one. I'll need to carefully analyze the structure and element names to determine the correct Send JSON Path.

upvoted 0 times

...

Stephaine

5 months ago

Hmm, I'm not confident. Is this about how companies might accidentally create environments that encourage unethical behavior?

upvoted 0 times

...

Wilburn

5 months ago

This question seems a bit tricky. I'll need to carefully read through the options to understand the nuances of Alibaba Cloud's intranet communication policies.

upvoted 0 times

...

Mable

5 months ago

The command looks straightforward, but I'm not 100% sure. I'll have to double-check the Docker documentation to be certain.

upvoted 0 times

...