Neural Network Deep Learning Architectures 2

Question 1

What problem does batch normalization primarily address in deep networks?

Accepted Answer

Internal covariate shift between layers

Answer

Overfitting on training data

Answer

Vanishing gradients only

Answer

Slow inference speed

Question 2

Which architecture introduced the concept of inception modules combining multiple filter sizes in parallel?

Accepted Answer

GoogLeNet (Inception)

Answer

VGG

Answer

AlexNet

Answer

ResNet

Question 3

What is the purpose of the bottleneck layer in an autoencoder?

Accepted Answer

To force the network to learn a compressed representation

Answer

To speed up training

Answer

To apply attention mechanisms

Answer

To perform classification

Question 4

In a U-Net architecture, what is the purpose of the skip connections between encoder and decoder?

Accepted Answer

To preserve fine-grained spatial information during reconstruction

Answer

To reduce computational cost

Answer

To apply dropout during decoding

Answer

To normalize batch statistics

Question 5

What distinguishes a capsule network from a standard CNN?

Accepted Answer

It encodes pose information as vectors rather than scalar activations

Answer

It uses recurrent connections

Answer

It applies self-attention to feature maps

Answer

It uses no pooling operations at all

Question 6

Which regularization technique randomly drops entire feature map channels during training in CNNs?

Accepted Answer

Spatial dropout

Answer

Weight decay

Answer

L1 regularization

Answer

Gradient clipping