Deep Learning Lab

What is the main advantage of using convolutional neural networks (CNNs)?

• They have lower numbers of layers

• They are more interpretable

• They can automatically detect important features

• They are faster to train

What is the purpose of the softmax function in neural networks?

• To compute the loss

• To convert logits to probabilities

• To initialize the weight

• To train the model

What is a neural network with multiple layers between the input and output layers called?

• Single Layer Perceptron

• Multi-layer Perceptron

• Convolutional Neural Network

• Recurrent Neural Network

What is the primary purpose of the "flatten layer" in a neural network?

• To add non-linearity to the model

• To perform weighted-sum

• To find the best weights and bias

• To convert multi-dimensional input into a one-dimensional vector

This figure represents the semantic diagram of the Convolutional neural network (CNN). The first part (Part1) of CNN performs .................................

• Classification

• Feature extraction

• Model Training

• Validation

This figure represents the semantic diagram of the Convolutional neural network (CNN). The second part (Part2) of CNN performs .................................

• Classification

• Feature extraction

• Model Training

• Validation

What is a commonly used ratio for splitting data into training, validation, and test sets in machine/deep learning?

• 50% training, 25% validation, 25% test

• 60% training, 30% validation, 10% test

• 70% training, 20% validation, 10% test

• 80% training, 10% validation, 10% test

What is the primary purpose of "optimization" in machine learning?

• To maximize the complexity of the model

• To minimize the number of training epochs

• To minimize the size of the model

• To find the best parameters that minimize the loss

In the context of deep learning, what does "training" refer to?

• Preparation of input data

• Adjusting the model's parameters using a training dataset to minimize a predefined loss function

• Evaluating the model's performance on a test dataset

• Deploying the trained model for making predictions

Which statement best describes the Rectified Linear Unit (ReLU) activation function?

• ReLU outputs values in the range [−1,1][-1, 1][−1,1]

• ReLU sets negative values to zero and retains positive values unchanged

• ReLU gives us the probability values

• ReLU is a variant of the sigmoid function