What is the primary purpose of using a validation set during neural network training?

To compute the final test accuracy

To train the model on additional data

To tune hyperparameters and monitor for overfitting

To initialize the weights of the network

A team is training a neural network on a small tabular dataset. The training accuracy reaches 99% while validation accuracy plateaus at 85%. The team needs to reduce overfitting without losing important feature information. Which approach is most appropriate?

Apply L2 regularization and dropout to the hidden layers

Use a larger batch size and a higher learning rate

Increase the number of hidden layers and neurons

Remove all regularization and train longer

A data scientist is training a neural network to predict house prices (in dollars) from features like square footage and location. The target variable is continuous and right-skewed. The model currently uses mean squared error (MSE) as the loss. The team notices that the model is heavily influenced by a few very expensive houses, causing poor predictions on the majority of homes. Which loss function change would best address this issue while still being appropriate for regression?

Switch to mean absolute error (MAE)

Switch to categorical cross-entropy

Keep MSE but add L2 regularization

GIAC Machine Learning Engineer

Domain 3Objective 3

Neural Networks GMLE Practice Questions (Page 3)

Part of the Advanced Machine Learning and Neural Networks domain, which makes up ~30% of our current practice bank. GIAC (SANS) does not publish an official question count, but from its 180-minute exam (~70–120 total, ~21–36 in this domain), expect 7–12 from this objective — we provide 39 practice questions to prepare you well beyond it. (estimate)

39questions here

8free pages

9concepts

Questions 11–15

11foundation · easy
What is the primary purpose of using a validation set during neural network training?
Select an answer first
12foundation · easy
Which hyperparameter controls the step size taken during gradient descent?
Select an answer first
13application · medium
A team is training a neural network on a small tabular dataset. The training accuracy reaches 99% while validation accuracy plateaus at 85%. The team needs to reduce overfitting without losing important feature information. Which approach is most appropriate?
Select an answer first
14application · medium
A data scientist is training a neural network to predict house prices (in dollars) from features like square footage and location. The target variable is continuous and right-skewed. The model currently uses mean squared error (MSE) as the loss. The team notices that the model is heavily influenced by a few very expensive houses, causing poor predictions on the majority of homes. Which loss function change would best address this issue while still being appropriate for regression?
Select an answer first
15application · medium
A machine learning engineer is training a neural network with a large number of features. The model performs well on training data but poorly on validation data. The engineer wants to reduce overfitting by shrinking the weights of less important features. Which regularization technique should be used?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to GMLE

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by GIAC (SANS). “GMLE” is a trademark of its owner, used for identification only.

Neural Networks GMLE Practice Questions (Page 3)

Questions 11–15

What is the primary purpose of using a validation set during neural network training?

Which hyperparameter controls the step size taken during gradient descent?

A team is training a neural network on a small tabular dataset. The training accuracy reaches 99% while validation accuracy plateaus at 85%. The team needs to reduce overfitting without losing important feature information. Which approach is most appropriate?

Keep studying — Advanced Machine Learning and Neural Networks