Amazon MLS-C01 Exam - Topic 5 Question 58 Discussion

Actual exam question for Amazon's MLS-C01 exam

Question #: 58
Topic #: 5

A data engineer at a bank is evaluating a new tabular dataset that includes customer dat

a. The data engineer will use the customer data to create a new model to predict customer behavior. After creating a correlation matrix for the variables, the data engineer notices that many of the 100 features are highly correlated with each other.

Which steps should the data engineer take to address this issue? (Choose two.)

AUse a linear-based algorithm to train the model.

BApply principal component analysis (PCA).

CRemove a portion of highly correlated features from the dataset.

DApply min-max feature scaling to the dataset.

EApply one-hot encoding category-based variables.

Show Suggested Answer

Suggested Answer: B, D

https://scikit-learn.org/stable/auto_examples/preprocessing/plot_all_scaling.html

by Novella at Jun 16, 2022, 05:52 AM

Limited Time Offer

25%

Off

Get Premium MLS-C01 Questions as Interactive Web-Based Practice Test or PDF

Contribute your Thoughts:

Submit Cancel

Roosevelt

4 months ago

PCA and feature removal are definitely the way to go here!

upvoted 0 times

...

Mattie

4 months ago

I thought one-hot encoding was for categorical data, not this?

upvoted 0 times

...

Carlton

4 months ago

Wait, why would you use a linear algorithm with high correlation?

upvoted 0 times

...

Moira

4 months ago

Totally agree, removing correlated features is key too.

upvoted 0 times

...

Britt

4 months ago

PCA is a solid choice for reducing dimensionality!

upvoted 0 times

...

Whitley

5 months ago

I don't think using a linear-based algorithm is the best choice here since it might not handle multicollinearity well.

upvoted 0 times

...

Lucille

5 months ago

I feel like we practiced a question similar to this, and I think using PCA was one of the recommended steps.

upvoted 0 times

...

Justine

5 months ago

I'm not entirely sure, but I think removing some of the correlated features could simplify the model. It sounds like a reasonable approach.

upvoted 0 times

...

Colene

5 months ago

I remember we discussed how PCA can help reduce dimensionality when features are highly correlated. That might be a good option here.

upvoted 0 times

...

Ezekiel

5 months ago

Hmm, I'm a bit unsure about the steps here. Do I need to create the regular file first, or can I do that last? And what's the deal with the vault password file - do I need to include that in the regular file somehow?

upvoted 0 times

...

Cordelia

5 months ago

This one seems pretty straightforward. I'm pretty sure the answer is B - Analogous estimate, since that's based on expert judgment from similar past projects.

upvoted 0 times

...

Lashandra

5 months ago

This is a tricky one, I'm not entirely sure. I'll need to review my notes on Cisco FMC high availability and synchronization to make an informed decision.

upvoted 0 times

...