MS-DS Master of Data science Deep Learning and Neural Networks 3

Question 1

Which deep learning architecture is best suited for sequential text classification tasks due to its ability to process inputs in order?

Accepted Answer

LSTM

Answer

CNN

Answer

Autoencoder

Answer

Feedforward network

Question 2

What is the 'vanishing gradient' problem in deep neural networks?

Accepted Answer

Gradients shrinking exponentially as they propagate back through many layers

Answer

Gradients becoming too large during backpropagation

Answer

Weights not updating due to zero learning rate

Answer

Loss oscillating around a local minimum

Question 3

Which technique involves training a neural network on augmented versions of training data (flips, crops, rotations) to improve generalization?

Accepted Answer

Data augmentation

Answer

Weight sharing

Answer

Knowledge distillation

Answer

Residual learning

Question 4

In deep learning, what does a residual (skip) connection accomplish in architectures like ResNet?

Accepted Answer

Allows gradients to flow directly through skip paths, easing optimization of very deep networks

Answer

Adds dropout between layers

Answer

Reduces the number of filters per layer

Answer

Applies batch normalization before each block

Question 5

What is the purpose of the softmax function in the output layer of a neural network for classification?

Accepted Answer

Convert raw logits into a probability distribution that sums to 1

Answer

Introduce sparsity in activations

Answer

Clip output values between -1 and 1

Answer

Normalize weights across all layers

Question 6

Which hyperparameter controls how much the model's weights are updated in response to the estimated error on each mini-batch?

Accepted Answer

Learning rate

Answer

Batch size

Answer

Number of epochs

Answer

Dropout rate