Deep Learning Quiz chap 3

What are the parameters related to a neural network that are learned during the model training?

• Model parameters

• Regularization

• Hyperparameters

• Optimization

What is bias in the context of a neural network?

• A measure of variability across models

• A measure of how far the predictions are from the true value

• A form of regularization

• A method to specify the number of training epochs

Which method is used to reduce fluctuations in the stochastic gradient descent algorithm?

• NAG - Nesterov Accelerated Gradient Descent

• Momentum Method

• AdaGrad - Adaptive Gradient Algorithm

• RMSProp - Root Mean Square Propagation

What does the Early Stopping method aim to prevent in neural network training?

• Gradient explosion

• Model convergence

• Underfitting

• Overfitting

What is the purpose of Weight Regularization in neural networks?

• To add noise to the training data

• To penalize large weights to prevent overfitting

• To reduce the learning rate

• To increase the number of epochs

What is the main responsibility of an optimizer in a neural network?

• To select the hyperparameters

• To change the learning rate and weights to minimize loss

• To define the activation function

• To perform data augmentation

What does the Batch Size refer to in neural network training?

• The learning rate

• The number of training instances in the batch

• The number of epochs

• The size of hidden units

Which optimization algorithm uses an exponential average of gradients as momentum?

• Adam

• Momentum Method

• AdaGrad

• RMSProp

What is the purpose of Data Augmentation in deep learning?

• To increase the size of the dataset

• To reduce the number of training epochs

• To increase the batch size

• To add noise to the training data

What is the objective of the Bayesian Optimization Algorithm?

• To explore the best hyperparameters efficiently

• To minimize the learning rate

• To maximize the number of training epochs

• To increase the batch size