Exam AWS Certified Machine Learning - Specialty topic 1 question 36 discussion

Exam question from Amazon's AWS Certified Machine Learning - Specialty

Question #: 36
Topic #: 1

[All AWS Certified Machine Learning - Specialty Questions]

A Machine Learning Specialist deployed a model that provides product recommendations on a company's website. Initially, the model was performing very well and resulted in customers buying more products on average. However, within the past few months, the Specialist has noticed that the effect of product recommendations has diminished and customers are starting to return to their original habits of spending less. The Specialist is unsure of what happened, as the model has not changed from its initial deployment over a year ago.
Which method should the Specialist try to improve model performance?

A. The model needs to be completely re-engineered because it is unable to handle product inventory changes.
B. The model's hyperparameters should be periodically updated to prevent drift.
C. The model should be periodically retrained from scratch using the original data while adding a regularization term to handle product inventory changes
D. The model should be periodically retrained using the original training data plus new data as product inventory changes.

Show Suggested Answer

Suggested Answer: D 🗳️

by rsimham at Dec. 9, 2019, 9:14 p.m.

Disclaimers:

- ExamTopics website is not related to, affiliated with, endorsed or authorized by Amazon.
- Trademarks, certification & product names are used for reference only and belong to Amazon.

Comments

Submit Cancel

rsimham

Highly Voted 3 years, 4 months ago

Ans: D

upvoted 27 times

...

JonSno

Most Recent 6 months ago

Selected Answer: D

The model performance degradation over time suggests concept drift—the relationship between input features and the target variable has changed. Since product recommendations depend on customer behavior, preferences, and product inventory, periodic retraining with updated data ensures the model adapts to these changes. Why Periodic Retraining? Customer preferences evolve: Buying patterns change over time due to seasons, trends, and external factors. New products get added, and old ones are discontinued: The model must learn about new items and stop recommending outdated ones. The dataset needs to reflect recent trends: Using new and historical data together ensures the model retains useful past knowledge while learning new patterns.

upvoted 2 times

...

james2033

11 months, 2 weeks ago

Selected Answer: D

'retrained using the original training data plus new data'

upvoted 1 times

...

VR10

1 year ago

I believe it should be B 1. The model performance has diminished gradually over the past few months, indicating the data distribution may have changed since initial deployment over a year ago. This is a classic sign of concept drift. 2. The model architecture and training procedure have remained unchanged since initial deployment. Updating the hyperparameters is a lighter approach than retraining the model from scratch, and can help prevent further performance deterioration if done periodically to adapt to changes in user preferences and product inventory.

upvoted 1 times

...

Valcilio

1 year, 11 months ago

Selected Answer: D

D is the answer!

upvoted 1 times

...

Peeking

2 years, 2 months ago

D is the answer. There has been a data drift resulting from new customer segment visiting the site. So, the model needs to be updated periodically with new data from the website.

upvoted 3 times

...