Amazon MLS-C01 Exam - Topic 2 Question 105 Discussion

Actual exam question for Amazon's MLS-C01 exam

Question #: 105
Topic #: 2

An ecommerce company wants to use machine learning (ML) to monitor fraudulent transactions on its website. The company is using Amazon SageMaker to research, train, deploy, and monitor the ML models.

The historical transactions data is in a .csv file that is stored in Amazon S3 The data contains features such as the user's IP address, navigation time, average time on each page, and the number of clicks for ....session. There is no label in the data to indicate if a transaction is anomalous.

Which models should the company use in combination to detect anomalous transactions? (Select TWO.)

AIP Insights

BK-nearest neighbors (k-NN)

CLinear learner with a logistic function

DRandom Cut Forest (RCF)

EXGBoost

Show Suggested Answer

Suggested Answer: D, E

To detect anomalous transactions, the company can use a combination of Random Cut Forest (RCF) and XGBoost models. RCF is an unsupervised algorithm that can detect outliers in the data by measuring the depth of each data point in a collection of random decision trees. XGBoost is a supervised algorithm that can learn from the labeled data points generated by RCF and classify them as normal or anomalous. RCF can also provide anomaly scores that can be used as features for XGBoost to improve the accuracy of the classification.References:

1: Amazon SageMaker Random Cut Forest

2: Amazon SageMaker XGBoost Algorithm

3: Anomaly Detection with Amazon SageMaker Random Cut Forest and Amazon SageMaker XGBoost

by Hubert at Sep 18, 2024, 11:19 PM

Limited Time Offer

25%

Off

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

Contribute your Thoughts:

Submit Cancel

Rozella

3 months ago

XGBoost seems too complex for just detecting anomalies, right?

upvoted 0 times

...

Celeste

3 months ago

Wait, no labels? How can they even train a model without them?

upvoted 0 times

...

Lavonda

3 months ago

Totally agree, RCF is great for this kind of task!

upvoted 0 times

...

Rene

4 months ago

K-nearest neighbors could work too, but it might struggle with high dimensions.

upvoted 0 times

...

Annabelle

4 months ago

I think Random Cut Forest is a solid choice for anomaly detection.

upvoted 0 times

...

Stefania

4 months ago

I think we should avoid models like XGBoost since they typically require labeled data. I’m leaning towards Random Cut Forest and maybe K-nearest neighbors, but I’m not completely sure.

upvoted 0 times

...

Trevor

4 months ago

I feel like we practiced a similar question, and I recall that Random Cut Forest was mentioned as a good choice for detecting anomalies. But I'm not confident about the second model.

upvoted 0 times

...

Shenika

4 months ago

I'm not entirely sure, but I think K-nearest neighbors might work too. It seems like it could help identify outliers based on the transaction features.

upvoted 0 times

...

Jolene

5 months ago

I remember we discussed using unsupervised models for anomaly detection since there's no label in the data. Maybe Random Cut Forest could be one of the options?

upvoted 0 times

...

Milly

5 months ago

Definitely going with IP Insights and Linear Learner. The IP address and navigation data should give us a good starting point to detect fraud, and the Linear Learner can help us model the anomalies.

upvoted 0 times

...

Robt

5 months ago

Hmm, I'm a bit unsure about this one. I'm considering k-NN and XGBoost, but I'm not sure if those are the best options since there's no labeled data. Might need to do some more research.

upvoted 0 times

...

Izetta

5 months ago

I think I'll go with IP Insights and Random Cut Forest for this one. The IP address and user behavior features seem like they could help identify anomalous transactions.

upvoted 0 times

...

Ira

5 months ago

I don't recall bankers' acceptances being tied to small firms. Trade credit seems like a safe guess!

upvoted 0 times

...

Tamekia

1 year ago

I believe Linear learner with a logistic function could also be helpful in detecting anomalous transactions. It's important to have a combination of models for better results.

upvoted 0 times

...

Filiberto

1 year ago

I agree with Lashawn, RCF is a good choice. But I also think they should consider using K-nearest neighbors (k-NN) for better accuracy.

upvoted 0 times

...

Laurene

1 year ago

I heard the ecommerce company is going to use a magical algorithm that can sniff out fraud like a bloodhound on a bacon trail. I call it the 'Sniff-Out-Scam-o-Matic 3000'!

upvoted 0 times

...

Lashawn

1 year ago

I think the company should use Random Cut Forest (RCF) to detect anomalous transactions.

upvoted 0 times

...

Rosamond

1 year ago

This is a piece of cake! IP Insights and Linear learner with a logistic function are the way to go. IP Insights for the IP address analysis, and the linear learner for the overall transaction patterns. Gotta catch those bad guys red-handed!

upvoted 0 times

Charlene

1 year ago

Let's deploy these models on Amazon SageMaker and monitor the results closely.

upvoted 0 times

...

Noble

1 year ago

Using both models in combination can enhance fraud detection capabilities.

upvoted 0 times

...

Fabiola

1 year ago

Linear learner with a logistic function can analyze transaction patterns effectively.

upvoted 0 times

...

Aliza

1 year ago

I agree, IP Insights can help identify suspicious IP addresses.

upvoted 0 times

...

Jaime

1 year ago

I'd recommend K-nearest neighbors (k-NN) and Random Cut Forest (RCF). k-NN is great for identifying outliers, and RCF can handle the lack of labeled data. Plus, it's always fun to watch the algorithm 'cut' through the fraudulent transactions!

upvoted 0 times

...

Mozell

1 year ago

Definitely go with IP Insights and XGBoost. XGBoost is a powerful algorithm that can handle the complex patterns in the data, and IP Insights will catch those sneaky fraudsters trying to hide behind proxy servers.

upvoted 0 times

Lashaunda

1 year ago

IP Insights will definitely help catch those fraudsters trying to hide their tracks.

upvoted 0 times

...

Titus

1 year ago

I agree, XGBoost is really powerful and can handle complex patterns well.

upvoted 0 times

...

Lynsey

1 year ago

IP Insights and XGBoost are great choices for detecting fraudulent transactions.

upvoted 0 times

...

Bambi

1 year ago

I think the company should use IP Insights and Random Cut Forest (RCF) to detect anomalous transactions. IP Insights can help identify suspicious IP addresses, while RCF can effectively handle the unsupervised nature of the data.

upvoted 0 times