Quiz 2

1. A convolutional layer has:

• 32 filters, each of size 5×5

• Input depth: 3

How many trainable parameters does this layer have (including biases)?

• 2400

• 2403

• 2410

• 2401

2. What is the primary purpose of the pooling layer in a CNN?

• A) Reducing the dimensionality and computational load.

• F) Combining feature maps from different layers.

• E) Increasing the size of the input images.

• G) Normalizing the pixel values of the input data.

3. In a CNN, if the filter size is 3x3, the stride is 1, and there is no padding, what happens to the size of the feature map?

• It fluctuates randomly.

• C) It decreases.

• It becomes larger.

• It remains unchanged.

4. What is the role of the activation function in a CNN?

• A) To normalize the input data.

• C) To reduce the number of parameters.

• D) To perform dimensionality reduction.

• B) To introduce non-linearity into the model.

5. Which of the following is a common activation function used in CNNs?

• Softmax

• Leaky ReLU

• Tanh

• C) ReLU

6. What does the term 'overfitting' refer to in the context of CNNs?

• The model is too complex and captures noise in the data.

• A) The model performs well on training data but poorly on unseen data.

• The model generalizes well to both training and unseen data.

• The model is trained on a diverse dataset with sufficient examples.

7. Which technique is commonly used to prevent overfitting in CNNs?

• Using dropout layers

• Implementing batch normalization

• B) Data augmentation

• Increasing the number of epochs

8. In CNNs, what is the purpose of using multiple filters in a convolutional layer?

• B) To capture different features from the input data.

• D) To simplify the activation function.

• C) To improve the accuracy of the model.

• A) To increase the size of the input data.

9. A CNN layer has:

• Input size: 32×32×16

• Filter size: 3×3, stride 1, padding 0

• 32 filters

How many activation values are computed in the output feature map?

• 32,768

• 16,384

• 49,152

• 64,000

10. An input image of size 64×64 is passed through a convolutional layer with the following settings:

• Filter size: 3×3

• Stride: 2

• Padding: 1

• Number of filters: 16

What will be the size of the output feature map?

• 32x32

• 64x64

• 33x33

• 35x35