CNN - Convolutional Neural Networks Overfitting and Regularization Questions and Answers

Question 1

A CNN model achieves 99% accuracy on the training dataset but only 75% on the validation dataset. Which phenomenon is occurring, and what is a common technique to address it?

Accepted Answer

Overfitting; apply Dropout to the fully connected layers.

Answer

Underfitting; increase model complexity by adding more layers.

Answer

Data leakage; remove features that are present in both training and validation sets.

Answer

Vanishing gradients; switch to a different activation function like Leaky ReLU.

Question 2

What is the primary effect of L1 regularization on the weights of a CNN?

Accepted Answer

It encourages some weights to become exactly zero, leading to a sparse model.

Answer

It encourages all weights to become smaller and more evenly distributed.

Answer

It has no direct effect on the weights but prunes entire network channels.

Answer

It normalizes the activations within each layer to have a mean of zero and a standard deviation of one.

Question 3

A data scientist is training a CNN on a small, specialized dataset of industrial parts. To prevent the model from memorizing the limited training examples, they decide to artificially expand the dataset. Which of the following techniques are they employing?

Accepted Answer

Data Augmentation

Answer

L2 Regularization

Answer

Early Stopping

Answer

Principal Component Analysis (PCA)

Question 4

Which of the following regularization techniques involves halting the training process when the model's performance on a validation set stops improving?

Accepted Answer

Early Stopping

Answer

Dropout

Answer

Weight Decay (L2)

Answer

Batch Normalization

Question 5

What is the primary goal of using regularization techniques like Dropout or L2 regularization when training a Convolutional Neural Network?

Accepted Answer

To reduce the model's complexity and improve its ability to generalize to unseen data.

Answer

To increase the speed of convergence during training.

Answer

To increase the model's performance on the training dataset.

Answer

To automatically determine the optimal number of convolutional layers.

Question 6

Which of the following is NOT a common technique used to combat overfitting in a CNN?

Accepted Answer

Increasing the number of epochs indefinitely.

Answer

Data Augmentation.

Answer

Dropout.

Answer

L2 Regularization (Weight Decay).

A CNN model achieves 99% accuracy on the training dataset but only 75% on the validation dataset.Which phenomenon is occurring, and what is a common technique to address it?

A CNN model achieves 99% accuracy on the training dataset but only 75% on the validation dataset.
Which phenomenon is occurring, and what is a common technique to address it?