Q: 10

A company trained an ML model on Amazon SageMaker to predict customer credit risk. The model shows 90% recall on training data and 40% recall on unseen testing data. Which conclusion can the company draw from these results?

Options

Discussion

Ryan Y. Oct 5, 2026 3:51 pm

A That recall gap between training and test is classic overfitting.

Karan Oct 10, 2026 7:32 am

Seen similar recall drop questions in the official guide, that's A. Practice exams cover this classic overfitting case too.

Jason N. Oct 3, 2026 2:04 am

A for sure. That huge gap in recall between train and test is typical overfitting. Doubt the issue is with data size.

Jordan L. Oct 9, 2026 10:47 pm

A , if the test set was tiny or non-representative, maybe D, but here it's textbook overfit.

Ryan J. Oct 11, 2026 7:53 am

Its A, high recall on train but low on test is textbook overfitting.

Casey F. Oct 11, 2026 4:14 am

A/B? If recall is high on training but way lower on test, that's usually classic overfitting (A). Underfitting (B) would show low recall on both, not just test. Unless I'm missing some context about data size, A fits better. But if training was super limited or data was weird, maybe B could pop up. Still think A is the main point here.

Morgan C. Oct 6, 2026 6:33 am

I don't think it's C. Big gap like 90% recall train vs 40% test usually means the model memorized training (A), not that there's not enough data. B's underfitting would show low recall across both, and D would need info about the test set being too small, which isn't in the question.

Liam M. Sep 28, 2026 10:53 pm

C or A? Hard to tell if they don't mention data size. Still, that recall gap usually means A.

Owen V. Oct 9, 2026 2:25 pm

C/D? Not really convinced by C, since a big recall drop like this is classic overfitting (A). D's possible only if test data is tiny, but nothing in the question says that. Pretty sure it's A but open to pushback.

QuinnF Sep 29, 2026 7:59 am

A . That kind of recall drop screams overfitting, not a data volume issue.

Be respectful. No spam.

Correct Answer:

Explanation

The model demonstrates high performance on the training data (90% recall) but significantly lower performance on the unseen testing data (40% recall). This large discrepancy is a classic symptom of overfitting. Overfitting occurs when a machine learning model learns the training data too well, including its noise and specific patterns, to the point where it negatively impacts the model's ability to generalize to new, unseen data. The model has effectively "memorized" the training set instead of learning the underlying, generalizable trends.

References

1. Amazon SageMaker Developer Guide: In the section on model tuning

it is stated that a sign of overfitting is when the model performs well on the training set but poorly on the validation set. "If the training accuracy is high and the validation accuracy is low

the model is likely overfitting to the training dataset."

Source: Amazon SageMaker Developer Guide

"Monitor and Analyze Training Jobs Using Metrics

" section on "Look for overfitting or underfitting".

2. Stanford University

CS229 Machine Learning Course Notes: The course materials define overfitting as a state where a model has a low training error but a high generalization (test) error. This discrepancy indicates the model has failed to capture the true underlying pattern and instead modeled the noise in the training data.

Source: Ng

A. (2023). CS229 Machine Learning Course Notes. Stanford University. Part V

"Learning Theory

" Section on Bias/variance tradeoff.

3. MIT OpenCourseWare

Introduction to Machine Learning: Lecture notes explain that overfitting happens when a model is too complex for the amount of training data available. This results in a model that fits the training data points almost perfectly but fails to generalize to new data points

leading to a large gap between training and test performance.

Source: MIT OpenCourseWare. (2020). 6.036 Introduction to Machine Learning

Fall 2020. Lecture 9 Notes: "Regularization

" Section 9.1.

A company trained an ML model on Amazon SageMaker to predict customer credit risk. The model shows 90% recall on training data and 40% recall on unseen testing data. Which conclusion can the company draw from these results?

Premium Access Includes