AIF-C01 Fundamentals of AI and ML Practice Question
During model training, the loss decreases rapidly for the first few epochs and then plateaus. The validation loss starts increasing after some epochs. What should the team do to improve generalization?
⚠ Common exam trap
The AIF-C01 exam often tests the misconception that overfitting is solved by increasing model complexity or training longer, when in fact the opposite is true—early stopping or regularization techniques are required to curb overfitting.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Early stopping
The validation loss increasing while training loss continues to decrease is a classic sign of overfitting. Early stopping (Option B) halts training when validation performance stops improving, preventing the model from memorizing noise in the training data and thereby improving generalization.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Increase learning rate
Why it's wrong here
A larger learning rate amplifies parameter updates, which can accelerate the divergence of validation loss once the model has passed its generalisation optimum. It is tempting because learning rate is a standard tuning parameter, and raising it would be correct when training loss decreases too slowly because updates are too small.
- ✓
Early stopping
Why this is correct
Rising validation loss while training loss plateaus signals overfitting beyond the optimal point. Early stopping halts training when validation loss stops improving, restoring the best weights and preventing further memorisation, thereby improving generalisation without changing the model architecture.
- ✗
Increase training epochs
Why it's wrong here
Training longer continues fitting the training set after validation loss has begun rising, deepening the overfitting the stem describes. It is tempting because more epochs usually improve accuracy, and extending training would be correct when both training and validation loss are still falling together, indicating the model has not yet converged.
- ✗
Add more layers
Why it's wrong here
Adding layers increases parameter count, giving the network more capacity to memorise training samples while validation loss already climbs. It is tempting because deeper networks often improve accuracy, and adding layers would be correct when training loss itself remains high, signalling the model lacks capacity to capture the underlying pattern.
Go deeper
Related to this question
About these practice questions
One of 862 original AIF-C01 practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This AIF-C01 practice question is part of Courseiva's free Amazon Web Services certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AIF-C01 exam.