P3

Question 1

Q: What is backpropagation?

Answer 1

A: An algorithm used for training neural networks by adjusting weights to minimise error.

Question 2

Q: Define natural language processing (NLP).

Answer 2

A: A field of AI focused on enabling machines to understand and respond to human language.

Question 3

Q: What is a dataset?

Answer 3

A: A collection of data used to train and evaluate machine learning models.

Question 4

Q: What is the purpose of a loss function?

Answer 4

A: To measure the difference between the predicted output and the actual target output.

Question 5

Q: What does a GPU do in machine learning?

Answer 5

A: It accelerates the processing of large-scale data and complex computations.

Question 6

​Q: What is an RNN?

Answer 6

A: Recurrent Neural Network, designed to handle sequential data.

Question 7

Q: Define LSTM.

Answer 7

A: Long Short-Term Memory, a type of RNN that handles long-term dependencies.

Question 8

Q: What is a transformer neural network?

Answer 8

A: A neural network using a self-attention mechanism for parallel processing of data.

Question 9

Q: What is BPTT?

Answer 9

A: Backpropagation through time, a variant of backpropagation for RNNs.

Question 10

Q: What is a memory cell state in LSTM?

Answer 10

A: It represents the information flowing through the network, managed by gates.

Question 11

Q: What is data cleaning?

Answer 11

A: The process of removing irrelevant, duplicate, or noisy data.

Question 12

Q: What is synthetic data?

Answer 12

A: Artificially generated data used to supplement real data.

Question 13

Q: What is bias in datasets?

Answer 13

A: Systematic errors that lead to unfair or discriminatory outcomes.

Question 14

Q: What is sampling bias?

Answer 14

A: When the dataset is not representative of the entire population.

Question 15

Q: What is selection bias?

Answer 15

A: Bias introduced when data is not randomly selected but chosen based on specific criteria

Question 16

​Q: Why is data privacy important?

Answer 16

A: To protect sensitive personal information from unauthorised access.

Question 17

Q: What is transparency in AI?

Answer 17

A: Making decision-making processes clear and understandable to users.

Question 18

Q: How can we prevent misinformation by chatbots?

Answer 18

A: By integrating fact-checking mechanisms.

Question 19

Q: What is accountability in chatbot ethics?

Answer 19

A: Determining who is responsible for the chatbot’s actions and decisions.

Question 20

Q: Define ethical use of chatbots.

Answer 20

A: Ensuring chatbots operate fairly, transparently, and responsibly, respecting user privacy.

Question 21

​Q: What is hyperparameter tuning?

Answer 21

A: The process of optimizing parameters that govern model training.

Question 22

Q: What is a self-attention mechanism?

Answer 22

A: A technique that captures relationships between words in a sequence by computing attention weights.

Question 23

Q: Define lexical analysis.

Answer 23

A: Breaking down text into individual words and sentences for further processing.

Question 24

Q: What is syntactical analysis?

Answer 24

A: Analsing the grammatical structure of a sentence to identify relationships between words.

Question 25

Q: What is semantic analysis?

Answer 25

A: Understanding the meaning of words and sentences beyond their surface structure.

Question 26

Q: What is model pruning?

Answer 26

A: Removing unnecessary neurons or connections in a neural network to reduce complexity.

Question 27

​Q: What is quantization in machine learning?

Answer 27

A: Reducing the precision of weights to lower bit sizes to enhance model efficiency.

Question 28

Q: Define knowledge distillation.

Answer 28

A: Transferring knowledge from a larger model to a smaller one to maintain performance while reducing complexity.

Question 29

Q: What is parallel processing?

Answer 29

A: Dividing tasks into smaller sub-tasks that can be processed simultaneously.

Question 30

Q: Why is cloud computing used in AI?

Answer 30

A: For scalable and flexible computing resources that can be adjusted based on demand.

Question 31

What is the primary function of a chatbot?

Answer 31

A: To provide automated responses to user queries using AI and NLP techniques

Question 32

Q: Define latency in chatbots.

Answer 32

A: The delay between a user's query and the chatbot’s response.

Question 33

Q: What is discourse integration?

Answer 33

A: Integrating the meaning of a sentence with the larger context of the conversation

Question 34

Q: What is pragmatic analysis?

Answer 34

A: Analysing the social, legal, and cultural context of a sentence to understand its intended meaning.

Question 35

Q: Why is contextual understanding important for chatbots

Answer 35

A: To provide coherent and relevant responses based on the broader conversation context.

Question 36

Q: What is historical bias?

Answer 36

A: Bias that occurs when training data reflects outdated patterns that may not be relevant to current scenarios.

Question 37

Q: What is labelling bias?

Answer 37

A: When the labels applied to training data are subjective, inaccurate, or incomplete.

Question 38

Q: What is linguistic bias?

Answer 38

A: Bias resulting from training data that favors certain dialects, vocabularies, or linguistic styles.

Question 39

Q: How can bias be detected in datasets?

Answer 39

A: By regularly auditing datasets and algorithms for biases and taking corrective actions.

Question 40

Q: Why is fairness important in chatbot interactions?

Answer 40

A: To ensure equitable service to all users, regardless of their background.

Question 41

​Q: What is a large language model (LLM)?

Answer 41

A: Advanced neural networks trained on vast amounts of text data to understand and generate human-like language.

Question 42

Q: Define natural language understanding (NLU).

Answer 42

A: A component of NLP focused on understanding the user’s input by analyzing linguistic features and context.

Question 43

Q: What is pre-processing in data handling?

Answer 43

A: Cleaning, transforming, and reducing data to improve its quality and make it suitable for training.

Question 44

Q: What is the vanishing gradient problem?

Answer 44

A: A problem in training deep neural networks where gradients become very small, making it difficult to update weights effectively.

Question 45

Q: What is a tensor processing unit (TPU)?

Answer 45

A: Custom hardware designed specifically to accelerate machine learning workloads.

Question 46

​Q: How does data augmentation help in training chatbots?

Answer 46

A: By generating additional data to increase the size and diversity of the dataset.

Question 47

Q: What is the role of encryption in data security?

Answer 47

A: To protect data both in transit and at rest from unauthorized access.

Question 48

Q: How can distributed computing benefit chatbots?

Answer 48

A: By parallelising processing across multiple machines to improve efficiency and reduce latency.

Question 49

Q: Why is user feedback important for chatbot improvement?

Answer 49

A: It helps identify and correct inaccuracies, continuously improving the chatbot’s performance.

Question 50

Q: What is explainable AI?

Answer 50

A: Techniques that make the decision-making process of AI systems understandable to users.