Probabilistic Models:

Question 1

What is a Bayesian Network?

Answer 1

A Bayesian Network is a graphical model that represents a set of variables and their conditional dependencies using directed acyclic graphs.

Question 2

True or False: In a Bayesian Network, nodes represent random variables.

Answer 2

True

Question 3

What does HMM stand for?

Answer 3

Hidden Markov Model

Question 4

Fill in the blank: In a Bayesian Network, the connections between nodes represent __________.

Answer 4

conditional dependencies

Question 5

What is the primary purpose of a Hidden Markov Model?

Answer 5

To model systems that are assumed to be Markov processes with unobserved (hidden) states.

Question 6

Multiple Choice: Which of the following is a characteristic of a Bayesian Network? A) Directed cycles B) Directed acyclic graph C) Unconditional independence D) None of the above

Answer 6

B) Directed acyclic graph

Question 7

True or False: HMMs can only be used for discrete state spaces.

Answer 7

False

Question 8

What is the role of the prior probability in a Bayesian Network?

Answer 8

The prior probability represents the initial beliefs about a variable before observing any evidence.

Question 9

Fill in the blank: The process of updating beliefs in a Bayesian Network upon receiving new evidence is called __________.

Answer 9

Bayesian inference

Question 10

What are the two main components of a Hidden Markov Model?

Answer 10

The states and the observations.

Question 11

Multiple Choice: In HMMs, the transition probabilities describe: A) The likelihood of moving from one state to another B) The likelihood of an observation given a state C) The initial state distribution D) All of the above

Answer 11

A) The likelihood of moving from one state to another

Question 12

True or False: In a Bayesian Network, the absence of an edge between two nodes implies independence.

Answer 12

True

Question 13

What is a joint probability distribution?

Answer 13

A joint probability distribution describes the probability of two or more random variables occurring simultaneously.

Question 14

Fill in the blank: The __________ algorithm is commonly used for training HMMs.

Answer 14

Baum-Welch

Question 15

What is the main advantage of using Bayesian Networks?

Answer 15

They allow for the representation of complex relationships among variables and enable probabilistic inference.

Question 16

Multiple Choice: Which of the following is NOT a component of a Bayesian Network? A) Nodes B) Edges C) States D) Conditional probability tables

Answer 16

C) States

Question 17

What does the term ‘Markov property’ refer to?

Answer 17

The Markov property refers to the principle that the future state of a process depends only on the current state and not on the sequence of events that preceded it.

Question 18

True or False: Bayesian Networks can only represent static relationships.

Answer 18

False

Question 19

What is the role of the emission probabilities in HMMs?

Answer 19

Emission probabilities define the likelihood of observing a particular output given a specific state.

Question 20

Fill in the blank: In a Bayesian Network, the __________ table provides the probabilities of a node given its parents.

Answer 20

conditional probability

Question 21

What is the difference between a prior and a posterior probability?

Answer 21

Prior probability is the initial belief before observing evidence, while posterior probability is the updated belief after taking evidence into account.

Question 22

Multiple Choice: Which algorithm is used for inference in Bayesian Networks? A) Viterbi algorithm B) Forward algorithm C) Variable elimination D) Expectation-Maximization

Answer 22

C) Variable elimination

Question 23

What is a causal relationship in the context of Bayesian Networks?

Answer 23

A causal relationship indicates that one variable directly affects another, represented by a directed edge in the network.

Question 24

True or False: HMMs can be used for speech recognition tasks.

Answer 24

True

Question 25

Fill in the blank: The __________ is used to find the most likely sequence of hidden states in an HMM.

Answer 25

Viterbi algorithm

Question 26

What is the purpose of the prior distribution in Bayesian inference?

Answer 26

To represent the initial beliefs about the parameters before observing the data.

Question 27

Multiple Choice: Which of the following techniques is commonly used for inference in HMMs? A) Particle filtering B) Gradient descent C) K-means clustering D) Linear regression

Answer 27

A) Particle filtering

Question 28

What is the significance of the conditional independence property in Bayesian Networks?

Answer 28

It simplifies the computation of joint probabilities and allows for efficient inference.

Question 29

True or False: HMMs can model systems with both observable and hidden states.

Answer 29

True

Question 30

What is the main limitation of Bayesian Networks?

Answer 30

They can become computationally expensive as the number of variables increases, leading to challenges in inference.

Question 31

What is a Vector Space Model?

Answer 31

A mathematical model for representing text documents as vectors in a continuous vector space.

Question 32

True or False: Word embeddings are a type of Vector Space Model.

Answer 32

True

Question 33

Fill in the blank: In a Vector Space Model, documents are represented as ______.

Answer 33

vectors

Question 34

What is the purpose of word embeddings?

Answer 34

To capture semantic meaning and relationships between words in a continuous vector space.

Question 35

Name one popular method for creating word embeddings.

Answer 35

Word2Vec

Question 36

What does the term 'ensemble model' refer to?

Answer 36

A machine learning model that combines predictions from multiple models to improve accuracy.

Question 37

Multiple Choice: Which of the following is an example of an ensemble model? A) Linear Regression B) Random Forest C) K-Means Clustering

Answer 37

B) Random Forest

Question 38

What is the main advantage of using ensemble models?

Answer 38

They reduce the risk of overfitting and improve predictive performance.

Question 39

True or False: Boosting is a technique that combines weak learners to create a strong learner.

Answer 39

True

Question 40

What is the difference between Bagging and Boosting?

Answer 40

Bagging reduces variance by averaging predictions, while Boosting reduces bias by combining weak learners sequentially.

Question 41

Fill in the blank: Random Forests are an ensemble method that uses ______ decision trees.

Answer 41

multiple

Question 42

What is the main idea behind the Random Forest algorithm?

Answer 42

To build multiple decision trees and aggregate their predictions for improved accuracy.

Question 43

Name one application of word embeddings.

Answer 43

Sentiment analysis

Question 44

Multiple Choice: Which of the following is NOT a characteristic of ensemble models? A) Reduces overfitting B) Combines multiple models C) Always increases computational cost

Answer 44

C) Always increases computational cost

Question 45

What is the role of the learning rate in boosting algorithms?

Answer 45

It controls how much each new model contributes to the ensemble.

Question 46

True or False: In a Vector Space Model, the dimensions of the space correspond to the number of unique words in the vocabulary.

Answer 46

True

Question 47

What is the purpose of dimensionality reduction in the context of word embeddings?

Answer 47

To reduce the number of features while preserving meaningful relationships between words.

Question 48

Fill in the blank: The technique of ______ uses a weighted average of weak learners to improve model performance in boosting.

Answer 48

weighted voting

Question 49

What is one disadvantage of using ensemble models?

Answer 49

Increased complexity and longer training times.

Question 50

Name one metric used to evaluate the performance of ensemble models.

Answer 50

Accuracy