Deep Learning

Question 1

In the backpropagation process, what does the term “gradient” refer to?

Answer 1

The slope of the loss function at a certain point

Question 2

In the context of backpropagation, what happens when a neural network is said to be “converging”?

Answer 2

The network is learning and the error is decreasing over iterations

Question 3

What is a common problem that can occur during backpropagation, especially with deep neural networks?

Answer 3

Gradient vanishing/exploding

Question 4

‘Underfitting’ in a machine learning model is a situation where:

Answer 4

The model has not learned the training data well enough

Question 5

What role does backpropagation play in addressing overfitting or underfitting in a neural network?

Answer 5

It helps to calculate gradients for updating weights, contributing to the learning process

Question 6

In unsupervised learning, what does the model learn from?

Answer 6

From only unlabeled data

Question 7

Which of the following is a common application of unsupervised learning?

Answer 7

Customer segmentation

Question 8

Which of the following is an example of a problem best suited for reinforcement learning?

Answer 8

Playing a game of chess

Question 9

Which learning method often involves the use of a ‘reward function’?

Answer 9

Reinforcement learning

Question 10

What is the role of a ‘label’ in supervised learning?

Answer 10

It is the desired output value for a given input data point

Question 11

What differentiates deep learning from traditional machine learning?

Answer 11

Deep learning algorithms are based on the structure and function of the human brain

Question 12

Why is the term ‘deep’ used in ‘deep learning’?

Answer 12

Because it refers to algorithms that use a deep stack of transformations from input to output

Question 13

What is the primary difference between traditional machine learning and deep learning?

Answer 13

Deep learning algorithms require more data

Question 14

How does the Transformer model, used in natural language processing tasks, handle sequential data?

Answer 14

It processes the sequence all at once using self-attention mechanisms

Question 15

Which of the following tasks can deep learning assist with in the field of natural language processing?

Answer 15

A. Spelling correction
B. Sentence generation
C. Machine translation
D. All of the above

Question 16

What is the role of word embeddings in natural language processing?

Answer 16

They are used to convert words into numerical vectors that capture semantic meaning

Question 17

Consider a scenario where you are analyzing tweets to determine overall sentiment (positive, negative, or neutral). Which type of sequence-to-sequence model configuration would be most appropriate for this task?

Answer 17

Many-to-One

Question 18

What is the main concept behind Generative Adversarial Networks (GANs)?

Answer 18

Two models, a generator and a discriminator, are trained together where the generator tries to fool the discriminator and the discriminator tries to not get fooled

Question 19

What is the Transformer model primarily used for?

Answer 19

Natural Language Processing (NLP)

Question 20

Which part of a GAN is responsible for generating new data?

Answer 20

The generator

Question 21

Which part of a GAN is trained to distinguish real data from fake data?

Answer 21

The discriminator

Question 22

What is one major advantage of Transformers over Recurrent Neural Networks (RNNs) for sequence prediction tasks?

Answer 22

Transformers can process longer sequences without forgetting earlier parts of the sequence

Question 23

What part of a neuron is similar to the activation function in a neural network?

Answer 23

Soma

Question 24

What type of artificial neural network architecture was inspired by the visual cortex?

Answer 24

Convolutional Neural Network (CNN)

Question 25

Which deep learning model is most similar to the way neurons in the human brain process sequential information over time?

Answer 25

Recurrent Neural Network (RNN)

Question 26

What is the equivalent of ‘neuroplasticity’ in the realm of artificial neural networks?

Answer 26

Transfer learning

Question 27

What is the role of ‘dendrites’ in a biological neuron and its equivalent in an artificial neuron?

Answer 27

Dendrites receive incoming signals from other neurons; equivalent to the input layer in an artificial neuron

Question 28

How is the concept of ‘firing’ in biological neurons represented in artificial neural networks?

Answer 28

By the activation function

Question 29

What is the primary difference between traditional artificial intelligence and deep learning?

Answer 29

Traditional AI uses rules and heuristics, while deep learning learns from data

Question 30

Which of the following is a popular deep learning framework?

Answer 30

TensorFlow

Question 31

What is a common problem encountered when training very deep neural networks?

Answer 31

Vanishing gradient problem

Question 32

What is the process of adjusting the weights of a neural network to minimize the difference between predicted and actual outputs called?

Answer 32

Backpropagation

Question 33

Which deep learning model architecture is used for generating new, realistic data samples from random noise?

Answer 33

Generative Adversarial Network (GAN)

Question 34

What is a limitation of deep learning compared to human learning?

Answer 34

Deep learning models cannot generalize knowledge beyond the trained data.

Question 35

What ethical concern arises with the use of deep learning algorithms in decision- making processes, such as in hiring or loan approval?

Answer 35

Bias and discrimination

Question 36

Which social issue is associated with the potential job displacement caused by automation and advancements in deep learning?

Answer 36

Income inequality

Question 37

What is the potential impact of deep learning on human interaction and social relationships?

Answer 37

Isolation and detachment from real-world interactions

Question 38

What is a concern related to the responsible use of deep learning models in autonomous systems, such as self-driving cars or military drones?

Answer 38

Lack of interpretability and explainability

Question 39

What is the potential impact of deep learning on education and learning systems?

Answer 39

Personalized learning experiences and adaptive tutoring

Question 40

Which human cognitive ability is critical for understanding and addressing biases in data and decision-making?

Answer 40

Ethical reasoning

Question 41

Which aspect of human communication is still challenging for deep learning models
to fully understand and interpret?

Answer 41

Sarcasm and humor

Question 42

What is the purpose of pre-training in the context of large language models like GPT?

Answer 42

To learn general language representations from a large corpus of text

Question 43

What is the primary advantage of using pre-trained large language models for natural language processing tasks?

Answer 43

They capture contextual and semantic information from large-scale data

Question 44

What is the purpose of the attention mechanism in Transformer-based language models like GPT?

Answer 44

To capture long-range dependencies and contextual information