Deep Learning Neural Network Fundamentals 3

Question 1

What is the softmax function used for in neural networks?

Accepted Answer

Converting raw scores into a probability distribution over classes

Answer

Introducing sparsity

Answer

Normalizing input features

Answer

Clipping gradient values

Question 2

What problem does batch normalization primarily address?

Accepted Answer

Internal covariate shift causing unstable training

Answer

Overfitting on training data

Answer

Exploding gradient values

Answer

Class imbalance in datasets

Question 3

In a neural network, what is an epoch?

Accepted Answer

One complete pass through the entire training dataset

Answer

A single weight update step

Answer

A subset of training data used in one update

Answer

The final evaluation step on test data

Question 4

Which technique randomly sets neuron outputs to zero during training to reduce overfitting?

Accepted Answer

Dropout

Answer

Batch normalization

Answer

L2 regularization

Answer

Weight decay

Question 5

What is the difference between stochastic gradient descent (SGD) and mini-batch gradient descent?

Accepted Answer

SGD updates weights using one sample; mini-batch uses a small subset

Answer

SGD uses all data per update; mini-batch uses one sample

Answer

SGD uses no learning rate; mini-batch does

Answer

They are identical algorithms

Question 6

What is a loss landscape in deep learning?

Accepted Answer

The high-dimensional surface of loss values across all possible weight configurations

Answer

A visualization of training accuracy over time

Answer

A plot of activation values per layer

Answer

The distribution of training labels