AI0-001 Machine Learning and Deep Learning Practice Question
A team is training a convolutional neural network (CNN) for medical image diagnosis. They have a limited dataset of 500 labeled images. Which strategy is most effective to improve model generalization?
⚠ Common exam trap
The AI0-001 exam often tests the misconception that increasing model complexity (depth or filters) always improves performance, but with limited data, the correct strategy is to use regularization techniques like data augmentation to combat 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
✓
Data augmentation
With only 500 labeled medical images, the primary challenge is overfitting due to limited data. Data augmentation (e.g., random rotations, flips, zooms) artificially expands the training set by creating varied but realistic transformations, which forces the CNN to learn invariant features and significantly improves generalization to unseen data.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Increasing network depth
Why it's wrong here
Increasing depth adds parameters, so a 500-image dataset is overfitted rather than generalised. It is tempting because deeper CNNs capture richer features on large datasets, but with limited labels the correct strategy is transfer learning or augmentation, which supply prior knowledge and variability instead of extra capacity.
- ✓
Data augmentation
Why this is correct
Data augmentation synthetically expands the 500 labelled images via rotations, flips and crops, directly addressing the limited-dataset constraint. This reduces overfitting and improves generalisation without requiring new labelled data, unlike transfer learning which needs a pretrained model or regularisation which only penalises complexity.
- ✗
Using a larger batch size
Why it's wrong here
A larger batch size changes gradient averaging per step; it does not add training signal, and with 500 images it can reduce generalisation. It is tempting because large batches speed training, but the effective remedy for scarce labelled data is augmentation or transfer learning, which increase data diversity.
- ✗
Reducing the number of filters
Why it's wrong here
Reducing filters lowers model capacity, which worsens underfitting on 500 images rather than improving generalisation. It is tempting as a regularisation-style simplification, but the correct approach for scarce labelled data is augmentation or transfer learning, which expand effective training signal without discarding representational power.
About these practice questions
Courseiva writes every AI0-001 question from scratch — 962 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This AI0-001 practice question is part of Courseiva's free CompTIA 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 AI0-001 exam.