AI0-001 AI Concepts and Foundations Practice Question
A data scientist is training a neural network to classify images of animals. The training accuracy is 99%, but validation accuracy is only 65%. Which technique should the data scientist use to address this issue?
⚠ Common exam trap
CompTIA often tests the distinction between techniques that improve training speed (batch normalization, learning rate tuning) versus those that improve generalization (dropout, regularization), and the trap here is that candidates may confuse overfitting with underfitting or assume that more training always helps.
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
✓
Add dropout layers to the network
The high training accuracy (99%) and low validation accuracy (65%) indicate overfitting, where the model memorizes the training data but fails to generalize. Adding dropout layers randomly drops neurons during training, which forces the network to learn more robust features and reduces overfitting. This technique is specifically designed to improve generalization without requiring more data or altering the learning rate.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Apply batch normalization
Why it's wrong here
Batch normalisation stabilises and speeds up training of deep networks but does not reduce the gap between training and validation accuracy. It is the right choice when training is unstable or slow to converge, not when the model is overfitting.
- ✗
Increase the number of training epochs
Why it's wrong here
Adding epochs worsens the existing train-validation gap, since the network already fits the training set at 99%; further passes deepen memorisation rather than improve generalisation. It is tempting because extra epochs help underfit models with high training loss, where both accuracies are still climbing together.
- ✓
Add dropout layers to the network
Why this is correct
Dropout layers randomly deactivate neurons during training, which curbs the network's reliance on particular features and reduces overfitting — the exact cause of the 99% training versus 65% validation gap. This directly targets the memorisation driving that disparity, improving generalisation to unseen images.
- ✗
Increase the learning rate
Why it's wrong here
Raising the learning rate makes optimisation less stable and typically widens the train-validation gap, not narrows it. It is tempting because a larger rate speeds convergence when training loss plateaus early, but here training accuracy is already 99%, so the bottleneck is generalisation.
About these practice questions
One of 962 original AI0-001 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 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.