Convolutional Neural Networks QUIZ (MCQ QUESTIONS AND ANSWERS)

Faster training times

Smaller model size

Improved performance on new tasks

Reduced memory usage

They reduce the size of feature maps

They introduce non-linearity into the network

They prevent overfitting by deactivating neurons

They increase the number of training epochs

The size of the convolutional kernel

The number of convolutional layers

The number of training epochs

The step size for moving the kernel across input data

Average pooling

Global max pooling

Max pooling

Min pooling

To increase the number of neurons in the network

To add noise to the input data

To reduce overfitting by randomly deactivating neurons during training

To speed up the training process

Increase model complexity

Improve computational efficiency

Preserve spatial dimensions

Enhance feature extraction

Introduce non-linearity

Compute gradients during backpropagation

Reduce spatial dimensions

Produce probability distributions

Convolutional layer

Fully connected layer

Pooling layer

Dropout layer

Introduce non-linearity

Extract features

Aggregate predictions

Reduce overfitting

Sigmoid

ReLU

Softmax

Tanh

Increase model complexity

Reduce spatial dimensions of feature maps

Apply activation functions

Perform convolution operations

To reduce the learning rate

To increase the complexity of the network

To stabilize training and mitigate vanishing/exploding gradients

To add regularization to the model

Convolutional layer

Pooling layer

Fully connected layer

Activation function layer

To prevent overfitting

To introduce non-linearity into the model

To preserve spatial dimensions and information at the edges of feature maps

To reduce the learning rate

They extract features from the input data

They reduce the spatial dimensions of feature maps

They aggregate learned features and make predictions

They apply non-linear transformations to feature maps

CNNs require less training data

CNNs can automatically detect important features from data

CNNs are faster to train

CNNs are simpler to implement

It increases the model's complexity

It reduces the number of training epochs required

It generates additional training samples by applying transformations to existing data

It reduces the learning rate

Sigmoid

Tanh

ReLU

Softmax

The output class labels

A small matrix used for convolution operations

The final layer of the network

The learning rate of the model

Convolutional layer

Activation function layer

Pooling layer

Fully connected layer

They perform fully connected operations on images

They require fewer parameters compared to fully connected layers

They add non-linearity to the network

They are less prone to overfitting

The number of times a filter is applied to the input data

The number of steps a kernel takes when moving across the input data during convolution

The depth of the feature maps produced by convolutional layers

The learning rate during training

Convolutional layer

Pooling layer

Fully connected layer

Output layer

To increase the complexity of the network

To add noise to the input data

To reduce overfitting by randomly deactivating neurons during training

To speed up training by skipping certain layers

Activation function layer

Pooling layer

Fully connected layer

Convolutional layer

Convolutional layer

Activation function layer

Pooling layer

Flatten layer

To initialize the model's weights

To reduce the learning rate

To accelerate training convergence and mitigate vanishing/exploding gradients

To add regularization to the model

Sigmoid

Tanh

ReLU (Rectified Linear Unit)

Softmax

To add noise to the input data

To prevent overfitting

To determine the learning rate

To preserve spatial dimensions and information at edges

Convolutional layer

Activation function layer

Pooling layer

Fully connected layer