Artificial Intelligence Neural Networks 2

Question 1

What distinguishes a convolutional neural network (CNN) from a standard feedforward network?

Accepted Answer

CNNs apply learnable filters that share weights spatially

Answer

CNNs use recurrent connections

Answer

CNNs have no hidden layers

Answer

CNNs use unsupervised learning only

Question 2

What is the purpose of a pooling layer in a CNN?

Accepted Answer

Reduce spatial dimensions and provide translational invariance

Answer

Add non-linearity

Answer

Increase the number of feature maps

Answer

Connect all neurons to the output

Question 3

What is batch normalization designed to do?

Accepted Answer

Normalize layer inputs to speed up and stabilize training

Answer

Increase the batch size during training

Answer

Apply dropout to each batch

Answer

Convert labels into one-hot encodings

Question 4

In a recurrent neural network (RNN), what makes it suitable for sequential data?

Accepted Answer

It maintains a hidden state that captures information from previous time steps

Answer

It uses convolutional filters

Answer

It processes all inputs in parallel

Answer

It uses attention mechanisms exclusively

Question 5

What problem do LSTMs (Long Short-Term Memory networks) solve compared to standard RNNs?

Accepted Answer

Vanishing gradients over long sequences

Answer

Slow inference speed

Answer

Lack of non-linear activations

Answer

High memory usage per parameter

Question 6

What is the softmax function used for in the output layer of a multi-class classifier?

Accepted Answer

Convert raw scores (logits) into a probability distribution over classes

Answer

Introduce sparsity in the weights

Answer

Normalize input features

Answer

Apply dropout to logits