exam questions

Exam DP-100 All Questions

View all questions & answers for the DP-100 exam

Exam DP-100 topic 5 question 32 discussion

Actual exam question from Microsoft's DP-100
Question #: 32
Topic #: 5
[All DP-100 Questions]

You create a binary classification model. You use the Fairlearn package to assess model fairness.
You must eliminate the need to retrain the model.
You need to implement the Fairlearn package.
Which algorithm should you use?

  • A. fairlearn.reductions.ExponentiatedGradient
  • B. fairlearn.postprocessing.ThresholdOptimizer
  • C. fairlearnpreprocessing.CorrelationRemover
  • D. fairlearn.reductions.GridSearch
Show Suggested Answer Hide Answer
Suggested Answer: B 🗳️

Comments

Chosen Answer:
This is a voting comment (?). It is better to Upvote an existing comment if you don't have anything to add.
Switch to a voting comment New
bbigwolf
Highly Voted 1 year, 8 months ago
Should be B
upvoted 8 times
...
cert_pz
Most Recent 5 months ago
Selected Answer: B
It states that the need to retrain the model should not apply, you can "enforce" fairness in a Model in 3 ways, before the training (preprocessing), during the training (inprocessing) and after the training (postprocessing). The Corrolation Remover is a preprocessing technique, therefore you would need to retrain the model. The correct answer is B.
upvoted 3 times
...
3than
5 months, 1 week ago
Nothing about fairlearn on MS Learn?
upvoted 1 times
...
PI_Team
7 months, 3 weeks ago
Selected Answer: B
The correct answer is B. fairlearn.postprocessing.ThresholdOptimizer. The ThresholdOptimizer algorithm in the Fairlearn package is a post-processing technique that you can use to adjust the threshold of a binary classification model’s predictions to improve fairness, without needing to retrain the model. This makes it a suitable choice given your requirement to eliminate the need to retrain the model. The other options listed are either preprocessing techniques or in-processing techniques which would require retraining of the model.
upvoted 2 times
...
Tommo565
1 year, 1 month ago
Selected Answer: B
As per other comments, B
upvoted 1 times
...
phdykd
1 year, 2 months ago
The ThresholdOptimizer algorithm allows you to adjust the decision threshold of the binary classification model to improve the balance between the accuracy of the model and the fairness of the model's predictions. This algorithm takes the original binary classification model as input and produces a new, fairer model without the need for retraining. Others require retraining the model.
upvoted 1 times
...
phdykd
1 year, 2 months ago
The correct sequence of actions is: B) Register each training file as a new datastore D) Add a new parameter in the module indicating the path to the training file E) Publish a training pipeline C) Run the training pipeline by using the studio portal Action A is not required as the path to the training file will be specified through the new parameter added in step 2.
upvoted 1 times
...
RamundiGR
1 year, 2 months ago
IT should be B
upvoted 1 times
...
BTAB
1 year, 3 months ago
Selected Answer: B
We want post processing, not pre processing
upvoted 1 times
...
michaelmorar
1 year, 3 months ago
Selected Answer: B
CorrelationRemover is a preprocessing tool -so that suggests you need to train again. I'll vote for B.
upvoted 1 times
...
Wayland
1 year, 7 months ago
B https://docs.microsoft.com/en-us/training/modules/detect-mitigate-unfairness-models-with-azure-machine-learning/4-mitigate-with-fairlearn
upvoted 3 times
...
Community vote distribution
A (35%)
C (25%)
B (20%)
Other
Most Voted
A voting comment increases the vote count for the chosen answer by one.

Upvoting a comment with a selected answer will also increase the vote count towards that answer by one. So if you see a comment that you already agree with, you can upvote it instead of posting a new comment.

SaveCancel
Loading ...
exam
Someone Bought Contributor Access for:
SY0-701
London, 1 minute ago