A deep CNN with batch normalization is training very slowly. The team notices that the loss decreases very slowly and the gradients are small. They suspect the learning rate is too low. They increase the learning rate from 0.0001 to 0.1. What is the most likely outcome?

Training may become unstable and the loss may diverge

Training will speed up proportionally without any issues

The gradients will become even smaller due to batch normalization

The model will converge to a better minimum because batch normalization allows arbitrary high learning rates

What is the purpose of the flattening step in a CNN before the fully connected layers?

To apply a nonlinear activation function to each feature map.

To reduce the number of parameters in the convolutional layers.

To increase the spatial resolution of the feature maps for better classification.

To convert the multi-dimensional feature maps into a one-dimensional vector that can be input to a fully connected layer.

In a CNN for image classification, what typically follows the flattening layer?

One or more fully connected layers

GIAC Machine Learning Engineer

Domain 3Objective 2

Convolutional Neural Networks GMLE Practice Questions (Page 4)

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 40 practice questions to prepare you well beyond it. (estimate)

40questions here

8free pages

10concepts

Questions 16–20

16expert · hard
A deep CNN with batch normalization is training very slowly. The team notices that the loss decreases very slowly and the gradients are small. They suspect the learning rate is too low. They increase the learning rate from 0.0001 to 0.1. What is the most likely outcome?
Select an answer first
17foundation · easy
What is the purpose of the flattening step in a CNN before the fully connected layers?
Select an answer first
18foundation · easy
In a CNN for image classification, what typically follows the flattening layer?
Select an answer first
19foundation · easy
Which technique is specifically designed to reduce overfitting by randomly dropping neurons during training?
Select an answer first
20foundation · easy
Which activation function is most commonly used in the hidden layers of a CNN to introduce nonlinearity?
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.

Convolutional Neural Networks GMLE Practice Questions (Page 4)

Questions 16–20

A deep CNN with batch normalization is training very slowly. The team notices that the loss decreases very slowly and the gradients are small. They suspect the learning rate is too low. They increase the learning rate from 0.0001 to 0.1. What is the most likely outcome?

What is the purpose of the flattening step in a CNN before the fully connected layers?

In a CNN for image classification, what typically follows the flattening layer?

Which technique is specifically designed to reduce overfitting by randomly dropping neurons during training?

Which activation function is most commonly used in the hidden layers of a CNN to introduce nonlinearity?

Keep studying — Advanced Machine Learning and Neural Networks