Unit I: Introduction to AI

1. What is Artificial Intelligence?

A. Making machines intelligent
B. Programming with logic
C. A system that thinks and acts like humans
D. All of the above
Answer: D
Explanation: AI involves building systems that think and act both rationally and like humans.

2. Which of the following is not an application of AI?

A. Speech recognition
B. Game playing
C. Weather forecasting
D. Robotics
Answer: C
Explanation: Weather forecasting uses statistical models; not all require AI.

3. Which is a foundational discipline of AI?

A. Philosophy
B. Psychology
C. Linguistics
D. All of the above
Answer: D
Explanation: AI draws from multiple disciplines to model intelligence.

4. Which year is considered the birth year of AI as a field?

A. 1945
B. 1950
C. 1956
D. 1969
Answer: C
Explanation: The Dartmouth Conference in 1956 marked the start of AI as a discipline.

5. Who is known as the father of AI?

A. Alan Turing
B. John McCarthy
C. Marvin Minsky
D. Arthur Samuel
Answer: B
Explanation: John McCarthy coined the term "Artificial Intelligence.

✅ Unit I: Introduction to Artificial Intelligence

Section A: Basics and Definitions

1. What is Artificial Intelligence?

A. Making computers intelligent
B. Programming with logic
C. A system that thinks and acts like humans
 D. All of the above
✅ Answer: D
📌 Explanation: AI involves creating systems that mimic human intelligence in thinking,
acting, and learning.

2. Which is a goal of AI?

A. Logical deduction
B. Knowledge representation
C. Learning
D. All of the above
✅ Answer: D

3. Which of the following is not a characteristic of AI?

A. Sensory perception
B. Mechanical strength
C. Problem-solving
D. Learning
✅ Answer: B

4. AI systems can be categorized into:

A. Weak AI and Strong AI
B. Narrow AI and Broad AI
C. Machine AI and Human AI
D. Real AI and Virtual AI
✅ Answer: A

5. What does "rational agent" mean in AI?

A. A system that acts randomly
B. A system that acts to achieve the best outcome
C. A system that thinks like a human
D. A human-controlled robot
✅ Answer: B

Section B: History of AI

6. When did the Dartmouth Conference take place?

A. 1943
B. 1950
C. 1956
D. 1965
✅ Answer: C

7. Who coined the term ‘Artificial Intelligence’?

A. Alan Turing
B. John McCarthy
C. Marvin Minsky
D. Arthur Samuel
✅ Answer: B
 8. Which of these is considered a milestone in early AI history?
A. ELIZA
B. Deep Blue
C. Watson
D. AlphaGo
✅ Answer: A

9. Alan Turing is famous for proposing:

A. Neural networks
B. Turing Test
C. AI ethics
D. Expert systems
✅ Answer: B

10. Which period saw the first AI Winter?

A. 1956-1960
B. 1974-1980
C. 1987-1993
D. 2000-2005
✅ Answer: B

Section C: Foundations of AI

11. Which of these is not a foundational discipline of AI?

A. Psychology
B. Philosophy
C. Geography
D. Linguistics
✅ Answer: C

12. AI is primarily concerned with:

A. Real-time OS
B. Intelligent behavior
C. Embedded systems
D. Network security
✅ Answer: B

13. Which of the following is a subfield of AI?

A. Networking
B. Natural Language Processing
C. Database systems
D. Operating Systems
✅ Answer: B

14. AI techniques involve:

A. Search algorithms
B. Knowledge representation
C. Machine learning
 D. All of the above
✅ Answer: D

15. Which mathematical concept is essential for AI?

A. Trigonometry
B. Calculus
C. Probability
D. Geometry
✅ Answer: C

Section D: AI Problems and Techniques

16. A well-defined AI problem should have:

A. Initial state
B. Goal state
C. Set of operations
D. All of the above
✅ Answer: D

17. Which of the following is an AI technique?

A. Divide and conquer
B. Search
C. Scheduling
D. Compiling
✅ Answer: B

18. Which is the first step in solving a problem using AI?

A. Implementation
B. Problem formulation
C. Algorithm writing
D. Testing
✅ Answer: B

19. What is the main goal of an AI agent?

A. To store information
B. To search and act rationally
C. To act randomly
D. To compile data
✅ Answer: B

20. A common AI problem-solving approach is:

A. Decision trees
B. Graph theory
C. Search in state-space
D. All of the above
✅ Answer: D

Section E: Applications of AI
 21. Which of the following is an application of AI?
A. Virtual assistants
B. Autonomous vehicles
C. Expert systems
D. All of the above
✅ Answer: D

22. AI is not typically used in:

A. Games
B. Weather forecasting
C. Medical diagnosis
D. Operating system development
✅ Answer: D

23. Which of the following uses AI for interaction?

A. ChatGPT
B. Amazon Alexa
C. Apple Siri
D. All of the above
✅ Answer: D

24. In which domain is AI used for fraud detection?

A. Agriculture
B. Education
C. Banking
D. Aviation
✅ Answer: C

25. Expert systems are designed to:

A. Repair machines
B. Simulate human expertise
C. Replace the Internet
D. Build new hardware
✅ Answer: B

Section F: Branches of AI

26. Which of these is a branch of AI?

A. Cybersecurity
B. Natural Language Processing
C. Circuit Design
D. Hardware Maintenance
✅ Answer: B

27. Which branch deals with enabling machines to ‘see’?

A. Robotics
B. Computer Vision
C. NLP
 D. Heuristics
✅ Answer: B

28. Machine Learning is best described as:

A. A programming language
B. A type of OS
C. A way to teach machines from data
D. A database tool
✅ Answer: C

29. Which of the following deals with human-machine interaction using voice or text?
A. NLP
B. Image Processing
C. Pattern Recognition
D. None of the above
✅ Answer: A

30. Which AI branch is used in autonomous vehicles?

A. NLP
B. Image recognition
C. Computer vision
D. All of the above
✅ Answer: D

Section G: Modern Tools and Platforms

31. Which of the following is a modern AI tool?

A. TensorFlow
B. Hadoop
C. MongoDB
D. Spark
✅ Answer: A

32. What is OpenAI known for?

A. Quantum computing
B. Developing GPT models
C. Database technologies
D. Cryptography
✅ Answer: B

33. ChatGPT is a model under which category?

A. Rule-based system
B. Generative AI
C. Reinforcement AI
D. Vision AI
✅ Answer: B

34. Which company created DeepMind?

A. Apple
B. Google
 C. Amazon
D. IBM
✅ Answer: B

35. IBM’s AI platform is called:

A. TensorAI
B. Alexa
C. Watson
D. Siri
✅ Answer: C

Section H: Miscellaneous

36. AI can help in:

A. Decision-making
B. Automation
C. Data analysis
D. All of the above
✅ Answer: D

37. Which of the following is an AI language?

A. Prolog
B. Assembly
C. C++
D. Pascal
✅ Answer: A

38. Which AI technique uses rules and logic?

A. Rule-based system
B. Neural Networks
C. Regression
D. Clustering
✅ Answer: A

39. Which field benefits from AI in terms of disease detection?

A. Healthcare
B. Agriculture
C. Education
D. Transport
✅ Answer: A

40. In AI, a heuristic is:

A. A rule of thumb
B. A programming language
C. A database
D. A robot
✅ Answer: A
Section I: Conceptual Scenarios

41. A self-driving car uses which combination of AI techniques?

A. Computer vision
B. NLP
C. Planning
D. All of the above
✅ Answer: D

42. Which AI domain is used in chatbots?

A. NLP
B. Machine Vision
C. Neural Networks
D. Game Theory
✅ Answer: A

43. The main input for machine learning is:

A. Data
B. Sound
C. Light
D. Rules
✅ Answer: A

44. In which year did IBM's Deep Blue beat the world chess champion?
A. 1990
B. 1997
C. 2001
D. 2005
✅ Answer: B

45. What does the Turing Test evaluate?

A. Memory power
B. Computational speed
C. Machine's intelligence
D. Network capability
✅ Answer: C

Section J: Review and Summary

46. Which is NOT a goal of AI?

A. Automation
B. Human-like thinking
C. Creating emotions in robots
D. Problem-solving
✅ Answer: C

47. Which of the following describes general AI?

A. AI that can perform one task
B. AI that simulates human intelligence in all tasks
C. AI that works in factories
 D. AI that fixes machines
✅ Answer: B

48. Intelligent agents perceive the environment using:

A. Effectors
B. Actuators
C. Sensors
D. Switches
✅ Answer: C

49. Which algorithm helps in finding optimal paths in AI problems?

A. Linear Regression
B. Depth-first search
C. A* Algorithm
D. Hashing
✅ Answer: C

50. AI problem-solving is often represented using:

A. Graphs
B. Networks
C. State-space trees
D. Linked lists
✅ Answer: C

Would you like me to continue with Unit II: Problem Spaces and Search next?

4o

You said:

yes

ChatGPT said:

Updated saved memory

Great! Let's begin with Unit I: Introduction to Artificial Intelligence. Here's a set of 50 MCQs with
solutions from Unit I:

✅ Unit I: Introduction to Artificial Intelligence – MCQs with Solutions

1–10: Basics and Definitions

1. What is the main goal of Artificial Intelligence?

a) To write programs that can solve problems
b) To create systems that can perform tasks requiring human intelligence
c) To develop programming languages
d) To replace human workers
Answer: b
2. Which of the following best defines Artificial Intelligence?
a) Automation of activities associated with human thinking
b) Making computers perform logical reasoning
c) Study of intelligent agents
d) All of the above
Answer: d

3. Which of the following is NOT a foundation of AI?

a) Philosophy
b) Biology
c) Chemistry
d) Psychology
Answer: c

4. The term “Artificial Intelligence” was coined by:

a) John McCarthy
b) Marvin Minsky
c) Alan Turing
d) Arthur Samuel
Answer: a

5. Which of these is a basic domain in AI?

a) Perception
b) Reasoning
c) Learning
d) All of the above
Answer: d

6. The Turing Test is designed to test a machine’s ability to:

a) Learn
b) Think logically
c) Exhibit intelligent behavior equivalent to a human
d) Speak naturally
Answer: c

7. Which of the following is a characteristic of AI problems?

a) Complex and ill-defined
b) Require symbolic processing
c) Require knowledge representation
d) All of the above
Answer: d

8. Which is not a type of AI technique?

a) Heuristic search
b) Knowledge representation
c) Algorithmic coding
d) Learning from experience
Answer: c

9. Which of these are applications of AI?

a) Robotics
 b) Expert Systems
c) Natural Language Processing
d) All of the above
Answer: d

10. Which is not a branch of AI?

a) Perception
b) Learning
c) Network Security
d) Reasoning
Answer: c

11–20: History and Foundations

11. The first AI program was developed in which year?

a) 1950
b) 1951
c) 1955
d) 1956
Answer: c

12. Which game was AI first able to play successfully?

a) Chess
b) Checkers
c) Tic-tac-toe
d) Poker
Answer: b

13. Which conference is considered the birth of AI?

a) Dartmouth Conference
b) MIT Symposium
c) Stanford Meeting
d) Carnegie Mellon Workshop
Answer: a

14. Which decade saw the first “AI winter”?

a) 1960s
b) 1970s
c) 1980s
d) 1990s
Answer: b

15. Which of these contributed significantly to AI foundations?

a) Neuroscience
b) Economics
c) Mathematics
d) All of the above
Answer: d

16. Strong AI aims to:

a) Act like humans
 b) Understand and think like humans
c) Perform tasks with brute force
d) Mimic behavior only
Answer: b

17. The field of AI was first formally introduced at:

a) MIT
b) Stanford
c) Dartmouth
d) Oxford
Answer: c

18. Who proposed the idea of a ‘thinking machine’?

a) Alan Turing
b) John McCarthy
c) Norbert Wiener
d) Ada Lovelace
Answer: a

19. Which of the following contributed to early neural network research?

a) McCulloch and Pitts
b) Turing
c) McCarthy
d) Nilsson
Answer: a

20. AI winter refers to:

a) A time of rapid development in AI
b) The birth of AI research
c) Periods of reduced AI funding and interest
d) Development of winter robots
Answer: c

I'll continue with MCQs 21–50 next (covering AI problems, techniques, applications, tools, and
branches). Would you like those now?

4o

You said:

i want now of unit 2

ChatGPT said:

Perfect! Here are 50 Multiple Choice Questions (MCQs) with solutions from Unit II: Problem Spaces
and Search.

✅ Unit II: Problem Spaces and Search – MCQs with Solutions

🔹 1–10: State Space, Production Systems & Problem Characteristics

1. In AI, a problem is defined as a:
a) Node and edge structure
b) Graph
c) State space
d) Logical rule
Answer: c

2. A problem space includes:

a) Initial state
b) Set of actions
c) Goal test
d) All of the above
Answer: d

3. What is a production system in AI?

a) A type of neural network
b) A system of if-then rules
c) A decision tree
d) A database
Answer: b

4. Which of these is a characteristic of a well-defined problem?

a) Known initial state
b) Well-defined goal state
c) Defined rules and constraints
d) All of the above
Answer: d

5. Production systems consist of:

a) A control strategy and rules
b) A database and sensors
c) Logical circuits
d) A list of heuristics
Answer: a

6. In a state space, the path from initial state to goal state is found using:
a) Decision trees
b) Search algorithms
c) Rule-based systems
d) Mathematical equations
Answer: b

7. Which of the following is a problem characteristic in AI?

a) Fully observable vs. partially observable
b) Deterministic vs. non-deterministic
c) Static vs. dynamic
d) All of the above
Answer: d

8. An AI system is said to be static if:

a) The environment changes frequently
 b) The environment remains fixed during search
c) The AI learns from feedback
d) The problem has no solution
Answer: b

9. In production systems, the set of rules is also known as:

a) Fact table
b) Knowledge base
c) Rule base
d) Command table
Answer: c

10. Which one is NOT an issue in designing search problems?

a) State representation
b) Defining success criteria
c) Input formatting
d) Choosing a control strategy
Answer: c

🔹 11–25: Uninformed Search Strategies

11. Which is an example of uninformed search?

a) Breadth First Search
b) A* algorithm
c) Best First Search
d) Hill Climbing
Answer: a

12. In BFS, nodes are expanded in order of:

a) Depth
b) Cost
c) Age
d) Level
Answer: d

13. Which search strategy uses a queue (FIFO)?

a) DFS
b) BFS
c) Hill Climbing
d) Simulated Annealing
Answer: b

14. Depth First Search (DFS) uses which data structure?

a) Queue
b) Heap
c) Stack
d) Priority Queue
Answer: c
15. Which is a limitation of DFS?
a) Space complexity
b) May not terminate
c) Always optimal
d) Slower than BFS
Answer: b

16. BFS always finds the:

a) Shortest path (in terms of cost)
b) Deepest path
c) Optimal solution (in terms of number of steps)
d) Most recent state
Answer: c

17. Which search explores both directions from start and goal?
a) BFS
b) DFS
c) Bi-directional Search
d) Greedy Search
Answer: c

18. Which search is a combination of BFS and DFS?

a) Simulated Annealing
b) A*
c) Iterative Deepening
d) Hill Climbing
Answer: c

19. Which search strategy reduces memory requirements compared to BFS?

a) Iterative Deepening
b) DFS
c) BFS
d) Bi-directional
Answer: a

20. Which search may get trapped in loops?

a) BFS
b) DFS
c) Iterative Deepening
d) A*
Answer: b

21. Uninformed search strategies do not use:

a) State space
b) Goal states
c) Heuristics
d) Successor functions
Answer: c

22. Which strategy expands shallowest unexpanded node first?

a) DFS
 b) BFS
c) Best First Search
d) Depth-Limited Search
Answer: b

23. Which of these is complete and optimal?

a) DFS
b) Hill Climbing
c) BFS
d) Simulated Annealing
Answer: c

24. Which search strategy expands deepest node first?

a) DFS
b) BFS
c) A*
d) Hill Climbing
Answer: a

25. Which strategy uses less memory but is not optimal?

a) BFS
b) DFS
c) A*
d) Best First Search
Answer: b

🔹 26–40: Informed Search Strategies

26. Which strategy uses a heuristic to guide the search?

a) DFS
b) BFS
c) A*
d) Depth-limited
Answer: c

27. What is a heuristic function?

a) Function that estimates cost to reach the goal
b) Function that checks validity
c) Function that defines a rule
d) None of the above
Answer: a

28. Generate and Test algorithm is a type of:

a) Blind search
b) Informed search
c) Constraint satisfaction
d) Production system
Answer: b
29. Hill climbing is similar to:
a) BFS
b) Greedy search
c) DFS
d) A*
Answer: b

30. Which of the following may suffer from local maxima?

a) BFS
b) DFS
c) Hill Climbing
d) A*
Answer: c

31. Simulated annealing differs from hill climbing in that:

a) It uses BFS
b) It accepts worse solutions with a probability
c) It requires no heuristic
d) It has no goal
Answer: b

32. Which search strategy uses both heuristic and path cost?
a) BFS
b) A*
c) Hill Climbing
d) DFS
Answer: b

33. What is the evaluation function of A search?*

a) g(n)
b) h(n)
c) f(n) = g(n) + h(n)
d) None
Answer: c

34. In Best First Search, nodes are ordered by:

a) Cost
b) Heuristic
c) Depth
d) Path length
Answer: b

35. Best First Search is also known as:

a) Greedy Search
b) A*
c) Simulated Annealing
d) Hill Climbing
Answer: a

36. Which heuristic search is both complete and optimal (with admissible heuristic)?
a) Hill Climbing
 b) Best First
c) A*
d) DFS
Answer: c

37. A heuristic is admissible if:

a) It overestimates the cost
b) It is always zero
c) It never overestimates the true cost
d) It guarantees shortest path
Answer: c

38. Constraint satisfaction problems are solved using:

a) DFS
b) Constraint propagation
c) Heuristics only
d) Iterative deepening
Answer: b

39. Which problem type fits into constraint satisfaction?

a) Sudoku
b) Maze navigation
c) Route finding
d) Chess
Answer: a

40. Which strategy uses randomness in the decision process?

a) A*
b) Simulated Annealing
c) BFS
d) Hill Climbing
Answer: b

🔹 41–50: Issues and Properties

41. Which is a major issue in designing search problems?

a) Choosing right state representation
b) Minimizing steps
c) Cost estimation
d) Goal generalization
Answer: a

42. State space can be defined as:

a) Set of possible configurations
b) Path to goal
c) Collection of rules
d) None
Answer: a
43. Search space complexity depends on:
a) Branching factor and depth
b) Node structure
c) Loop detection
d) Data type
Answer: a

44. Which is NOT true about heuristic search?

a) May reduce search time
b) Uses problem-specific knowledge
c) Guarantees optimal solution
d) Can be incomplete
Answer: c

45. In local search, we use:

a) State space only
b) Paths
c) Trees
d) Networks
Answer: a

46. Local maxima, ridges, and plateaus are problems in:

a) A*
b) Simulated Annealing
c) Hill Climbing
d) BFS
Answer: c

47. The completeness of a search algorithm refers to:

a) Memory usage
b) Finding goal if it exists
c) Speed
d) Optimality
Answer: b

48. The optimality of a search algorithm refers to:

a) Finding the goal
b) Using minimal memory
c) Finding least cost path
d) Depth-first search
Answer: c

49. Which is both complete and optimal (among uninformed)?

a) DFS
b) BFS
c) Hill Climbing
d) Greedy Search
Answer: b

50. Which method improves Hill Climbing by avoiding local maxima?

a) A*
 b) Best First
c) Simulated Annealing
d) DFS
Answer: c

UNIT III: KNOWLEDGE REPRESENTATION & STATISTICAL REASONING

1–15: Approaches and Issues in Knowledge Representation

1. Which of the following is not a knowledge representation technique?

A) Semantic Networks
B) Frames
C) Neural Networks
D) Production Rules
Answer: C) Neural Networks

2. The process of converting raw facts into meaningful data is called:

A) Interpretation
B) Reasoning
C) Representation
D) Abstraction
Answer: D) Abstraction

3. Which representation is best suited for inheritance?

A) Propositional Logic
B) Frames
C) Production Rules
D) Semantic Nets
Answer: B) Frames

4. A drawback of propositional logic is:

A) Expresses complex relationships
B) Cannot represent variables
C) It is ambiguous
D) It can represent time
Answer: B) Cannot represent variables

5. In semantic networks, knowledge is represented as:

A) Production rules
B) Graphs
C) Tables
D) Trees
Answer: B) Graphs

6. Which of the following is not a component of a frame?

A) Slots
B) Fillers
C) Nodes
D) Values
Answer: C) Nodes
7. In knowledge representation, "default reasoning" refers to:
A) Always true facts
B) Inference based on general assumptions
C) Deductive logic
D) Statistical calculation
Answer: B) Inference based on general assumptions

8. A knowledge base that contains facts and rules is called:

A) Procedural knowledge
B) Declarative knowledge
C) Strategic knowledge
D) Structural knowledge
Answer: B) Declarative knowledge

9. Which logic allows representation of relations and quantifiers?

A) Propositional logic
B) Predicate logic
C) Modal logic
D) Temporal logic
Answer: B) Predicate logic

10. Which is a major issue in knowledge representation?

A) Syntax
B) Reasoning
C) Knowledge acquisition
D) Truth value
Answer: C) Knowledge acquisition

11. Frames are similar to:

A) Classes in OOP
B) Hashmaps
C) Databases
D) Arrays
Answer: A) Classes in OOP

12. Which method can represent uncertainty in knowledge?

A) Predicate Logic
B) Certainty Factors
C) Semantic Net
D) Truth Tables
Answer: B) Certainty Factors

13. In Propositional Logic, a proposition is:

A) Always true
B) A declarative statement
C) A predicate
D) A question
Answer: B) A declarative statement

14. Which is not an issue in knowledge representation?

A) Expressiveness
 B) Efficiency
C) Syntax highlighting
D) Inference
Answer: C) Syntax highlighting

15. The predicate logic formula ∀x (Bird(x) → CanFly(x)) means:

A) All birds can fly
B) Some birds can fly
C) Birds don’t fly
D) x can fly
Answer: A) All birds can fly

16–30: Propositional & Predicate Logic, Reasoning

16. Propositional logic deals with:

A) Objects and their properties
B) Logical connectives between propositions
C) Uncertainty
D) Probabilities
Answer: B) Logical connectives between propositions

17. The backward reasoning process starts from:

A) Data
B) Conclusion
C) Rules
D) Evidence
Answer: B) Conclusion

18. Modus ponens is a rule in:

A) Predicate logic
B) Semantic logic
C) Propositional logic
D) Bayesian logic
Answer: C) Propositional logic

19. "If it rains, then the ground is wet." This is an example of:
A) Conjunction
B) Implication
C) Negation
D) Biconditional
Answer: B) Implication

20. Forward chaining starts with:

A) Hypotheses
B) Goals
C) Facts
D) Proof
Answer: C) Facts
21. Which reasoning method is used in expert systems for diagnosis?
A) Forward chaining
B) Backward chaining
C) Brute force
D) Random guessing
Answer: B) Backward chaining

22. In propositional logic, ¬(A ∨ B) is equivalent to:

A) ¬A ∧ ¬B
B) A ∧ B
C) ¬A ∨ ¬B
D) A ∨ B
Answer: A) ¬A ∧ ¬B (De Morgan’s Law)

23. Predicate logic is more powerful than propositional logic because:

A) It uses symbols
B) It can model functions
C) It can express relationships between objects
D) It uses truth tables
Answer: C) It can express relationships between objects

24. Quantifier "∃x" denotes:

A) All x
B) Not x
C) Exists an x
D) No x
Answer: C) Exists an x

25. Which logic can represent complex relationships like "Father of Father is Grandfather"?
A) Propositional logic
B) Predicate logic
C) Fuzzy logic
D) Probability logic
Answer: B) Predicate logic

26. Logical inference is the process of deriving new statements using:

A) Sensing
B) Learning
C) Deduction
D) Compilation
Answer: C) Deduction

27. In backward reasoning, the search proceeds:

A) Data to goal
B) Goal to data
C) Randomly
D) In cycles
Answer: B) Goal to data

28. If both a statement and its negation are true, the logic system is said to be:
A) Inconsistent
 B) Complete
C) Sound
D) Coherent
Answer: A) Inconsistent

29. Predicate logic uses which type of variables?

A) Boolean
B) Quantified
C) Local
D) Indexed
Answer: B) Quantified

30. Which logical operator is represented by ∧ ?

A) OR
B) NOT
C) AND
D) IMPLIES
Answer: C) AND

31–50: Statistical Reasoning, Probabilistic Logic & Bayesian Networks

31. Which theorem relates prior and posterior probability?

A) Fuzzy Theorem
B) Dempster-Shafer Theorem
C) Bayes’ Theorem
D) Zadeh's Theorem
Answer: C) Bayes’ Theorem

32. In Bayesian networks, nodes represent:

A) Facts
B) Variables
C) Rules
D) Inference paths
Answer: B) Variables

33. A belief network is also called:

A) Genetic network
B) Decision tree
C) Bayesian network
D) Semantic web
Answer: C) Bayesian network

34. Certainty factors are used in:

A) Predicate logic
B) Rule-based expert systems
C) State space search
D) Neural networks
Answer: B) Rule-based expert systems
35. What does P(A|B) denote in probability theory?
A) Probability of A and B
B) Probability of B given A
C) Probability of A given B
D) Independent event
Answer: C) Probability of A given B

36. Which statistical reasoning model allows combining multiple evidence sources?
A) Fuzzy logic
B) Dempster-Shafer theory
C) Propositional logic
D) Forward chaining
Answer: B) Dempster-Shafer theory

37. Bayesian networks work best with:

A) Complete data
B) Incomplete data
C) No data
D) Logical reasoning only
Answer: A) Complete data

38. Certainty Factor (CF) ranges from:

A) -1 to 1
B) 0 to 1
C) 0 to 100
D) -100 to 100
Answer: A) -1 to 1

39. Which is not a component of Bayesian network?

A) Directed Acyclic Graph
B) Conditional Probability Table
C) Inference Engine
D) Nodes
Answer: C) Inference Engine

40. The principle of maximum entropy is used to:

A) Minimize uncertainty
B) Maximize unknowns
C) Create fuzzy logic
D) Reduce entropy
Answer: A) Minimize uncertainty

41. Bayes’ theorem is applicable in case of:

A) Independent events only
B) Dependent events
C) Discrete logic
D) Fuzzy rules
Answer: B) Dependent events

42. The output of a Bayesian network is:

A) Rule
 B) Conclusion
C) Posterior Probability
D) Certainty Factor
Answer: C) Posterior Probability

43. In Dempster-Shafer theory, belief functions combine:

A) Facts and rules
B) Probabilities
C) Evidence
D) Frames
Answer: C) Evidence

44. Which method helps in reasoning under uncertainty?

A) Predicate logic
B) Production rules
C) Statistical reasoning
D) Forward chaining
Answer: C) Statistical reasoning

45. Which of these allows updating beliefs dynamically?

A) Semantic nets
B) Production rules
C) Bayesian networks
D) Propositional logic
Answer: C) Bayesian networks

46. Bayesian inference is best described as:

A) Deductive reasoning
B) Abductive reasoning
C) Probabilistic reasoning
D) Statistical logic
Answer: C) Probabilistic reasoning

47. Which theory supports belief intervals instead of single probability values?
A) Bayes’ theorem
B) Certainty factor model
C) Dempster-Shafer theory
D) Predicate logic
Answer: C) Dempster-Shafer theory

48. Which is used in decision support systems for uncertainty modeling?

A) Decision trees
B) Predicate logic
C) Bayesian networks
D) Resolution refutation
Answer: C) Bayesian networks

49. A probability distribution that represents the whole system in a Bayesian network is called:
A) Joint probability distribution
B) Prior distribution
C) Evidence function
 D) Knowledge base
Answer: A) Joint probability distribution

50. Certainty Factor of +1 indicates:

A) Complete disbelief
B) Complete belief
C) Neutral
D) Contradiction
Answer: B) Complete belief

🔹 Unit IV: Game Playing and Natural Language Processing - 50 MCQs

🎮 Game Playing (Q1–Q25)

1. What is the primary goal of game playing in AI?

A. Increase computation
B. Simulate graphics
C. Make optimal decisions
D. Store large data
→ C. Make optimal decisions

2. Which of the following games is considered a perfect information game?

A. Poker
B. Chess
C. Bridge
D. Scrabble
→ B. Chess

3. The evaluation function is used in which of the following?

A. Heuristic search
B. Game playing
C. Knowledge representation
D. NLP
→ B. Game playing

4. What is the purpose of the Minimax algorithm?

A. To sort data
B. To minimize loss
C. To optimize moves in two-player games
D. To learn user preferences
→ C. To optimize moves in two-player games

5. Which of the following is NOT a property of the minimax algorithm?

A. Complete
B. Optimal
C. Non-deterministic
D. Deterministic
→ C. Non-deterministic
6. Which condition applies for pruning in Alpha-Beta pruning?
A. Beta ≤ Alpha
B. Alpha = Beta
C. Alpha ≥ 0
D. Beta < 0
→ A. Beta ≤ Alpha

7. Which method improves Minimax performance?

A. Simulated annealing
B. Alpha-beta pruning
C. Genetic algorithm
D. Expectimax
→ B. Alpha-beta pruning

8. The Minimax algorithm assumes:

A. Random opponent
B. Optimal opponent
C. Cooperative opponent
D. Learning opponent
→ B. Optimal opponent

9. Which of the following is a terminal state in game playing?

A. Intermediate board state
B. Start of the game
C. A final win/loss/draw state
D. Random move
→ C. A final win/loss/draw state

10. In a zero-sum game, the gain of one player is:

A. Equal to the gain of the other
B. Double the gain of the other
C. Equal to the loss of the other
D. Irrelevant
→ C. Equal to the loss of the other

11. How does alpha-beta pruning help?

A. Increases depth
B. Avoids repeated states
C. Eliminates unnecessary branches
D. Uses more memory
→ C. Eliminates unnecessary branches

12. What is a drawback of Minimax without pruning?

A. Incomplete search
B. Slow for large trees
C. Non-optimal moves
D. No solution
→ B. Slow for large trees

13. What does alpha represent in alpha-beta pruning?

A. Best value for minimizing player
 B. Worst value for maximizing player
C. Best value for maximizing player
D. Random threshold
→ C. Best value for maximizing player

14. What is a ply in game trees?

A. One level of game tree
B. A complete solution
C. A backward move
D. A memory buffer
→ A. One level of game tree

15. Games with random elements (like dice) are handled by:
A. Minimax
B. Expectiminimax
C. BFS
D. DFS
→ B. Expectiminimax

16. The best-case time complexity of Alpha-Beta pruning is:

A. O(b^d)
B. O(b^(d/2))
C. O(d^b)
D. O(b^2)
→ B. O(b^(d/2))

17. Which of the following is a two-player game?

A. Sudoku
B. Solitaire
C. Tic-Tac-Toe
D. Crossword
→ C. Tic-Tac-Toe

18. The branching factor in chess is approximately:

A. 2
B. 10
C. 35
D. 100
→ C. 35

19. Which algorithm explores all children nodes before going deeper?
A. DFS
B. BFS
C. Minimax
D. Hill Climbing
→ B. BFS

20. A good evaluation function should be:

A. Slow
B. Irrelevant
C. Accurate and fast
 D. Complex
→ C. Accurate and fast

21. Which of the following increases efficiency in game playing?

A. Redundant branching
B. Repetition
C. Pruning
D. Random choices
→ C. Pruning

22. Game tree size increases:

A. Linearly
B. Exponentially
C. Logarithmically
D. Constantly
→ B. Exponentially

23. The time taken for game search increases with:

A. Branching factor
B. Evaluation speed
C. Memory size
D. Sound quality
→ A. Branching factor

24. Which search technique is best for a two-player perfect information game?
A. BFS
B. DFS
C. A*
D. Minimax
→ D. Minimax

25. Which component determines the winner in a game tree?

A. Utility function
B. State function
C. Constraint
D. Initial node
→ A. Utility function

Natural Language Processing (Q26–Q50)

26. What does NLP stand for?

A. Natural Logic Processing
B. Natural Language Processing
C. Neural Language Parser
D. None of the above
→ B. Natural Language Processing

27. The first step in NLP is:

A. Syntax analysis
B. Lexical analysis
 C. Semantic analysis
D. Pragmatic analysis
→ B. Lexical analysis

28. Which model is used in spelling correction?

A. A*
B. Soundex
C. Genetic
D. BFS
→ B. Soundex

29. Tokenization is a part of which phase?

A. Morphological analysis
B. Syntax analysis
C. Parsing
D. Pragmatic analysis
→ A. Morphological analysis

30. A parse tree represents:

A. Speech signals
B. Sentence structure
C. Semantics
D. Bag of words
→ B. Sentence structure

31. The Bag of Words model ignores:

A. Word meaning
B. Word order
C. Word frequency
D. All of these
→ B. Word order

32. Which of the following is a voice-based NLP assistant?

A. Google Docs
B. Siri
C. GitHub
D. VS Code
→ B. Siri

33. What is POS tagging?

A. Position of text
B. Parsing over sentences
C. Part-of-speech tagging
D. None of the above
→ C. Part-of-speech tagging

34. Chatbots are trained using:

A. Structured data
B. Decision trees
C. NLP techniques
 D. A* algorithm
→ C. NLP techniques

35. Which is not an NLP tool?

A. SpaCy
B. NLTK
C. Siri
D. MySQL
→ D. MySQL

36. Ambiguity in NLP is resolved using:

A. Spelling check
B. Context and grammar
C. Keywords
D. None
→ B. Context and grammar

37. Stemming involves:

A. Removing stop words
B. Reducing words to their base form
C. Splitting sentences
D. Predicting next word
→ B. Reducing words to their base form

38. Which NLP technique checks grammar?

A. Semantic analysis
B. Syntax analysis
C. Pragmatic analysis
D. Lexical analysis
→ B. Syntax analysis

39. Which algorithm is used in keyword search and matching phonetically?

A. Soundex
B. A*
C. KNN
D. DFS
→ A. Soundex

40. Named Entity Recognition (NER) extracts:

A. Numbers
B. Named entities like person, location
C. Grammar
D. Synonyms
→ B. Named entities like person, location

41. Which is the most challenging task in NLP?

A. Tokenization
B. Part-of-speech tagging
C. Ambiguity resolution
D. Word count
→ C. Ambiguity resolution
42. Voice assistants are examples of:
A. Supervised ML
B. NLP Applications
C. Data Mining
D. Fuzzy systems
→ B. NLP Applications

43. Which assistant uses NLP and ML?

A. Siri
B. Google Assistant
C. Alexa
D. All of the above
→ D. All of the above

44. Parse tree construction involves:

A. Parsing input grammar
B. Counting words
C. Splitting documents
D. Matching sound
→ A. Parsing input grammar

45. Phonetics-based NLP algorithm:

A. POS Tagging
B. Soundex
C. NER
D. BERT
→ B. Soundex

46. The key goal of NLP is to make computers:

A. Recognize images
B. Understand human language
C. Translate programming languages
D. Play games
→ B. Understand human language

47. A chatbot simulates:

A. Voice generation
B. Search engine
C. Human conversation
D. None
→ C. Human conversation

48. Which component maps speech to text?

A. ASR
B. NER
C. OCR
D. NLP
→ A. ASR (Automatic Speech Recognition)

49. In NLP, semantics deals with:

A. Meaning of words and sentences
 B. Word count
C. Parsing
D. Translation
→ A. Meaning of words and sentences

50. NLP is used in:

A. Text classification
B. Machine translation
C. Speech recognition
D. All of the above
→ D. All of the above

🔹 Unit V: Generative AI & Prompt Engineering – 50 MCQs with Solutions

Generative AI – Fundamentals

1. What does Generative AI primarily focus on?

A) Analyzing existing data
B) Generating new content
C) Sorting data
D) Fetching information
Answer: B

2. Which of the following is an example of Generative AI?

A) Excel
B) Google Search
C) ChatGPT
D) Calculator
Answer: C

3. Generative AI uses which core technique?

A) Rule-based logic
B) Neural Networks
C) Sorting algorithms
D) Hash tables
Answer: B

4. What is the main difference between traditional AI and Generative AI?

A) Generative AI can't learn
B) Traditional AI creates content
C) Generative AI generates new data
D) Traditional AI is cloud-based
Answer: C

5. Which of these is NOT a type of Generative AI model?

A) GAN
B) VAEs
C) CNN
D) Diffusion Models
Answer: C
 6. Generative Adversarial Networks (GANs) consist of:
A) A single model
B) Generator and Discriminator
C) Parser and Transformer
D) Only Generator
Answer: B

7. Diffusion Models are known for:

A) Text generation
B) Noise-based image generation
C) Data sorting
D) Numerical analysis
Answer: B

8. Which architecture powers most modern GenAI tools like ChatGPT?

A) LSTM
B) RNN
C) Transformer
D) CNN
Answer: C

9. Generative AI tools learn from:

A) Structured tables only
B) Only mathematical formulas
C) Large datasets of various types
D) Human memories
Answer: C

10. What is "text-to-image" in GenAI?

A) Data encryption
B) Image labeling
C) Generating images from descriptions
D) Screenshot capturing
Answer: C

Prompt Engineering – Fundamentals

11. What is Prompt Engineering?

A) Writing code
B) Fine-tuning hyperparameters
C) Crafting inputs to guide AI output
D) Reducing model size
Answer: C

12. Prompt Engineering is mostly associated with:

A) Procedural languages
B) Database management
C) Large Language Models
 D) Operating systems
Answer: C

13. Which of these is a good example of a prompt?

A) "Hello!"
B) "Translate this sentence into French: How are you?"
C) "Run the program"
D) "Save file"
Answer: B

14. A well-crafted prompt can:

A) Confuse the model
B) Reduce output quality
C) Improve accuracy and relevance
D) Slow the model down
Answer: C

15. Few-shot prompting provides:

A) No examples
B) One example
C) Multiple examples
D) Poor output
Answer: C

16. Zero-shot prompting provides:

A) Many examples
B) No examples
C) Labels only
D) Fine-tuned models
Answer: B

17. What is the key goal of prompt engineering?

A) Make the model faster
B) Make prompts longer
C) Get accurate, context-rich output
D) Write JavaScript
Answer: C

18. Which prompt is better for summarization?

A) “Run the code.”
B) “Summarize the following paragraph in one sentence.”
C) “Calculate the sum.”
D) “Describe output.”
Answer: B

19. Prompt tuning involves:

A) Adjusting hardware
B) Finetuning neural weights
C) Optimizing input prompts
D) Upgrading browsers
Answer: C
 20. Chain-of-thought prompting is used for:
A) Image labeling
B) Sequential reasoning
C) Noisy data
D) Data fetching
Answer: B

Language Models & Prompt Patterns

21. GPT stands for:

A) Generative Predictive Transformer
B) Generative Pretrained Transformer
C) General Processing Tool
D) Generating Precise Text
Answer: B

22. In LLMs, temperature parameter controls:

A) Speed
B) Creativity/randomness
C) File size
D) Output length
Answer: B

23. Which of the following is NOT a prompt pattern?

A) Instruction
B) Input-output pair
C) Command line
D) Q&A
Answer: C

24. The "Persona" prompt pattern means:

A) Give a random answer
B) Assume a character or role
C) Repeat previous output
D) Translate language
Answer: B

25. To get factual answers from GPT models, use:

A) Complex riddles
B) Vague instructions
C) Clear, precise prompts
D) Random keywords
Answer: C

26. What is a key element of a good prompt?

A) Ambiguity
B) Vagueness
C) Clarity and specificity
 D) Use of slang
Answer: C

27. Which element helps in multi-step reasoning prompts?

A) Humor
B) Chain-of-thought
C) Grammar errors
D) Capital letters
Answer: B

28. A generative AI model that creates images from text:

A) GPT
B) DALL·E
C) Siri
D) Cortana
Answer: B

29. Prompt tuning is best described as:

A) Software update
B) Model training
C) Automatic refinement of input prompts
D) A hardware optimization
Answer: C

30. Which of these improves LLM output?

A) Short prompts only
B) Random phrases
C) Context-rich prompts
D) Capital letters only
Answer: C

Applications and Use Cases

31. Which of these uses prompt engineering?

A) Text-to-speech converters
B) Large Language Models
C) Barcode readers
D) Thermometers
Answer: B

32. Generative AI can generate:

A) Noise only
B) Only structured data
C) Text, images, and audio
D) Only text
Answer: C

33. Which is NOT a use case of Generative AI?

A) Content creation
B) Fraud detection
 C) Creative design
D) Language translation
Answer: B

34. Siri and Alexa use:

A) Static scripts only
B) Natural Language Processing
C) Only hardware interfaces
D) Image recognition
Answer: B

35. Prompt Engineering is critical in:

A) Rule-based AI
B) Image compression
C) Large Language Models
D) Excel macros
Answer: C

36. Which is an AI assistant developed by Amazon?

A) Cortana
B) Siri
C) Alexa
D) Google Now
Answer: C

37. What is the main goal of prompt engineering in AI chatbots?

A) Increase cost
B) Reduce intelligence
C) Improve output relevance
D) Decrease power usage
Answer: C

38. ChatGPT uses prompts to:

A) Store memory
B) Generate specific responses
C) Encrypt data
D) Parse images
Answer: B

39. Prompt Engineering helps reduce:

A) Accuracy
B) Model training
C) Output errors
D) User interface design
Answer: C

40. Soundex algorithm is used in NLP for:

A) Image detection
B) Spell correction
C) Phonetic matching
 D) Translation
Answer: C

Prompt Techniques and Real-World Tips

41. Prompt chaining refers to:

A) Running multiple AI models
B) Using multiple prompts in sequence
C) Coding in Python
D) Compressing prompts
Answer: B

42. Which helps in classifying GenAI model types?

A) Output type (text/image/audio)
B) File format
C) Zip code
D) OS version
Answer: A

43. What does a low temperature value (e.g., 0.2) do in GPT?

A) Increases randomness
B) Generates unpredictable output
C) Makes output more deterministic
D) Stops generation
Answer: C

44. A "prompt template" is useful for:

A) Making GUIs
B) Reusing prompt patterns
C) Encrypting messages
D) Generating charts
Answer: B

45. In LLMs, larger context window allows:

A) Shorter inputs
B) More data retention
C) Lower performance
D) Frequent resets
Answer: B

46. Which is a bad practice in prompt design?

A) Including examples
B) Using clear instructions
C) Being vague
D) Using grammar
Answer: C

47. Prompt engineering is a blend of:

A) Hardware and software
B) UI and UX
 C) Psychology and communication
D) Math and robotics
Answer: C

48. Good prompt engineering leads to:

A) Incoherent responses
B) Better AI interpretability
C) Data deletion
D) Slower models
Answer: B

49. "Temperature" in LLM controls:

A) Electricity use
B) Prompt length
C) Creativity of responses
D) Training speed
Answer: C

50. Prompt patterns are used to:

A) Style the output
B) Design interface
C) Structure inputs for better results
D) Tune GPUs
Answer: C

✅ Unit VI – MCQs

📌 Section A: ChatGPT Advanced Data Analysis

1. Which feature distinguishes ChatGPT Advanced Data Analysis from regular ChatGPT?
A) Grammar correction
B) Ability to perform code execution and data analysis
C) Sentiment analysis
D) Access to weather data
Answer: B
Explanation: ChatGPT Advanced Data Analysis can run Python code and perform complex
data manipulations.

2. What kind of files can ChatGPT Advanced Data Analysis handle for automation?
A) .exe
B) .mp4
C) .zip
D) .dll
Answer: C
Explanation: It can open and process contents from zip files.

3. What is the primary use of ChatGPT in data visualization?

A) Typing documents
B) Designing UX
C) Creating graphs using code
 D) Video editing
Answer: C

4. Which tool is typically used by ChatGPT Advanced Data Analysis for generating plots?
A) Excel
B) PowerPoint
C) Matplotlib
D) Notepad
Answer: C

5. Structured data refers to:

A) Video files
B) Image datasets
C) Organized data in tabular format
D) PDF documents
Answer: C

6. What does ChatGPT Advanced Data Analysis rely on to work with documents?
A) File size
B) File extension only
C) Text extraction and parsing
D) Image resolution
Answer: C

7. Which of the following is NOT supported by ChatGPT Advanced Data Analysis?

A) SQL query generation
B) Web scraping with live access
C) Data cleaning with code
D) CSV file interpretation
Answer: B

8. One important consideration in Human-AI planning is:

A) Data redundancy
B) File format compatibility
C) Role assignment and collaboration
D) High CPU usage
Answer: C

9. What is the key factor in effective error identification in AI systems?

A) Data encryption
B) Use of neural networks
C) Log analysis and debugging
D) GPU usage
Answer: C

10. Which media type can ChatGPT Advanced Data Analysis process?
A) JPEG
B) CSV
C) MP3
D) TXT
Answer: A
📌 Section B: Advanced Topics in Artificial Intelligence

11. Which of the following is a common type of machine learning?

A) Supervised learning
B) Audio learning
C) Visual learning
D) Cognitive learning
Answer: A
Explanation: Supervised learning is the most common and involves training a model on
labeled data.

12. What is the primary goal of machine learning?

A) To simulate human behavior
B) To create deterministic algorithms
C) To find patterns in data and make predictions
D) To generate random data
Answer: C

13. Which of the following is NOT a type of machine learning?

A) Supervised learning
B) Reinforcement learning
C) Unsupervised learning
D) Experimental learning
Answer: D

14. What is the primary function of a neural network?

A) Encrypting data
B) Performing basic mathematical operations
C) Mimicking the human brain to solve problems
D) Generating random values
Answer: C

15. Which of the following is an example of a neural network?

A) Convolutional Neural Network (CNN)
B) Decision Tree
C) Linear Regression
D) K-Nearest Neighbors
Answer: A

16. In a neural network, which component is responsible for adjusting weights during training?
A) Output layer
B) Hidden layer
C) Backpropagation algorithm
D) Activation function
Answer: C

17. Which type of algorithm is used to model the process of natural evolution in Genetic
Algorithms?
A) Clustering
 B) Mutation and Crossover
C) Gradient descent
D) Nearest neighbor search
Answer: B

18. Which of the following best describes Genetic Algorithms?

A) Optimization algorithms based on the concept of natural selection
B) Algorithms used for supervised learning
C) Algorithms for clustering unstructured data
D) Algorithms for feature extraction
Answer: A

19. Which of the following is true for fuzzy logic systems?

A) Fuzzy logic only deals with binary outcomes.
B) Fuzzy logic is an extension of traditional Boolean logic.
C) Fuzzy logic can only handle crisp values.
D) Fuzzy logic is not applicable in machine learning.
Answer: B
Explanation: Fuzzy logic allows for degrees of truth, unlike binary true/false logic.

20. Which of the following is a key feature of fuzzy logic?

A) Precision
B) Inability to handle uncertainty
C) Ability to handle uncertainty and imprecision
D) Exclusivity in decision making
Answer: C

21. What type of machine learning technique does reinforcement learning use?
A) Labeling data
B) Trial and error with feedback
C) Clustering without labels
D) Optimizing data with known outputs
Answer: B

22. In which scenario would you typically use a Genetic Algorithm?

A) Data prediction with labeled datasets
B) Optimization problems like scheduling and resource allocation
C) Clustering large datasets
D) Time series forecasting
Answer: B

23. What does a "fitness function" in a Genetic Algorithm do?

A) Measures the probability of mutation
B) Evaluates the quality of solutions
C) Decides which genes to eliminate
D) Calculates the convergence speed
Answer: B

24. In fuzzy logic, which of the following is NOT a typical operation?

A) Fuzzy AND
B) Fuzzy OR
 C) Fuzzy XOR
D) Exact match
Answer: D

25. Which technique is commonly used to make neural networks more efficient?
A) Regularization
B) Linear regression
C) Data scaling
D) Feature elimination
Answer: A

26. What is overfitting in machine learning?

A) The model performs well on unseen data but poorly on training data.
B) The model performs too well on training data, capturing noise.
C) The model generalizes well to all data.
D) The model is underperforming on both training and test data.
Answer: B

27. In machine learning, the term "feature" refers to:

A) A result predicted by the model
B) A transformation function
C) An input variable used to train the model
D) A hyperparameter
Answer: C

28. Which of the following is a common application of neural networks?

A) Image classification
B) Data visualization
C) Sentiment analysis
D) Both A and C
Answer: D

29. What does the "activation function" in a neural network determine?

A) How weights are adjusted during training
B) Whether a neuron fires or not
C) The output of the network
D) The number of neurons in a layer
Answer: B

30. What is the primary difference between supervised and unsupervised learning?
A) Supervised learning uses labeled data, while unsupervised learning uses unlabeled data.
B) Unsupervised learning requires human intervention, while supervised does not.
C) Supervised learning only works with small datasets.
D) Unsupervised learning is used for predictions, while supervised is for classifications.
Answer: A

31. Which of the following algorithms is best suited for regression tasks?
A) K-means
B) Linear regression
C) Decision trees
 D) Support vector machine
Answer: B

32. What is an important benefit of using a convolutional neural network (CNN) for image
processing?
A) It works well on sequential data.
B) It is good at capturing hierarchical spatial features.
C) It is optimized for text processing.
D) It needs fewer data preprocessing steps.
Answer: B

33. Which of the following is NOT a characteristic of deep learning models?

A) They require large amounts of labeled data.
B) They can automatically extract features from raw data.
C) They are shallow networks with fewer layers.
D) They perform well on image and speech recognition tasks.
Answer: C

34. What type of problem is best solved by a decision tree algorithm?

A) Image recognition
B) Classification or regression problems
C) Time series forecasting
D) Natural language processing
Answer: B

35. Which of the following is a common use of reinforcement learning?

A) Video game AI
B) Predictive modeling
C) Text classification
D) Data clustering
Answer: A

36. What does the term "ensemble method" refer to in machine learning?
A) Using multiple models together to improve performance
B) Creating a single complex model
C) Generating random values to train a model
D) Selecting a subset of features for training
Answer: A

37. Which of the following methods helps mitigate overfitting in machine learning models?
A) Increasing the training data size
B) Using more complex models
C) Decreasing the number of features
D) Adding noise to the data
Answer: A

38. Which of the following is NOT a feature of a convolutional layer in a CNN?

A) Filters (kernels)
B) Pooling
C) Fully connected neurons
 D) Batch normalization
Answer: C

39. In machine learning, "gradient descent" is used to:

A) Find the optimal model parameters
B) Reduce the number of features in a dataset
C) Automatically label training data
D) Convert continuous data into categorical data
Answer: A

40. Which algorithm is primarily used for time series forecasting?

A) K-means clustering
B) ARIMA
C) Naive Bayes
D) Random Forest
Answer: B

41. In the context of AI, what does NLP stand for?

A) Non-linear programming
B) Natural Language Processing
C) Numerical Linear Programming
D) Network Learning Protocol
Answer: B

42. Which of the following is an example of a rule-based system in AI?

A) Chatbot using decision trees
B) Expert systems for medical diagnosis
C) Neural networks for image classification
D) Reinforcement learning for game playing
Answer: B

43. The term "backpropagation" refers to:

A) Optimizing the neural network by adjusting weights based on the error
B) Sending information to the output layer
C) Splitting data into training and testing sets
D) Applying dropout to the network
Answer: A

44. Which is the primary goal of using fuzzy logic in control systems?
A) To model binary decisions
B) To handle uncertainty and imprecision in decision-making
C) To replace neural networks
D) To simplify the process of feature extraction
Answer: B

45. Which technique is commonly used in Natural Language Processing for text vectorization?
A) Bag of Words
B) Linear Regression
C) K-means clustering
D) Convolutional networks
Answer: A
 46. In supervised learning, the data used to train the model is:
A) Unlabeled
B) Labeled
C) Structured
D) Cleaned
Answer: B

47. Which AI technique is most commonly used for pattern recognition?

A) Neural networks
B) Genetic algorithms
C) Decision trees
D) Bayesian networks
Answer: A

48. Which machine learning algorithm is best suited for handling large datasets with many
features?
A) K-nearest neighbors
B) Support vector machines
C) Decision trees
D) Random Forests
Answer: D

49. In a neural network, what is the function of the output layer?

A) To perform feature extraction
B) To classify or make predictions
C) To adjust the weights during training
D) To add noise for robustness
Answer: B

50. Which concept in AI is essential for understanding reinforcement learning?

A) Decision trees
B) Exploration and exploitation
C) Neural network architecture
D) Clustering data points
Answer: B

✅ Unit I – AI Fundamentals (Additional MCQs)

51. What is the main purpose of artificial intelligence in real-world applications?

A) To replicate human behavior
B) To perform calculations
C) To replace humans in all jobs
D) To solve problems that require human-like thinking
Answer: D

52. Which of the following is NOT considered an AI problem?

A) Image recognition
B) Natural language understanding
C) Sorting numbers
D) Autonomous driving
Answer: C
 53. Which of the following is the best example of a decision-making system in AI?
A) A facial recognition system
B) A recommendation system
C) An AI chatbot that answers questions
D) An AI system that plays chess
Answer: B

54. Which technique is used in AI to solve complex problems by simulating human reasoning?
A) Expert systems
B) K-means clustering
C) Nearest neighbor algorithms
D) Linear regression
Answer: A

✅ Unit II – Problem Spaces and Search (Additional MCQs)

55. In AI, a state space search problem typically includes all of the following EXCEPT:
A) Initial state
B) Goal state
C) Transition model
D) Evaluation function
Answer: D

56. A production system consists of all the following components EXCEPT:

A) A knowledge base
B) A set of production rules
C) A method for evaluating the goal state
D) A search algorithm
Answer: C

57. Which of the following search strategies would be most appropriate for finding a solution
to a problem where you want to explore all possible options systematically?
A) Depth-first search
B) Breadth-first search
C) Hill climbing
D) Best-first search
Answer: B

✅ Unit III – Knowledge Representation and Statistical Reasoning (Additional MCQs)

58. Which of the following is true about predicate logic in AI?

A) It only deals with facts and propositions
B) It is used to represent knowledge about objects, properties, and relationships
C) It does not support reasoning or inference
D) It is only used for mathematical calculations
Answer: B
 59. Which of the following is a major issue with knowledge representation in AI?
A) Efficient storage of data
B) Lack of high-quality data
C) Representing ambiguity and uncertainty
D) Unclear objective functions
Answer: C

60. Bayesian networks are particularly useful for representing:

A) Certainty factors
B) Probability distributions and dependencies among variables
C) Linguistic structures
D) Temporal sequences of events
Answer: B

✅ Unit IV – Game Playing and Natural Language Processing (Additional MCQs)

61. In the Min-Max algorithm for game playing, which of the following is true?
A) It evaluates the current game state based on possible moves for both players
B) It only looks at the next possible move
C) It does not consider the opposing player's strategy
D) It is applicable only to two-player, zero-sum games
Answer: A

62. Which of the following is NOT a phase of Natural Language Processing (NLP)?
A) Tokenization
B) Parsing
C) Classification
D) Proofreading
Answer: D

63. Which of the following algorithms is used for spell checking in NLP?
A) Naive Bayes
B) Soundex algorithm
C) K-means clustering
D) PageRank
Answer: B

✅ Unit V – Generative AI and Prompt Engineering (Additional MCQs)

64. Generative AI models like GPT are typically used to:

A) Predict future outcomes from historical data
B) Generate new data that resembles the input data
C) Perform classification tasks
D) Cluster data into groups
Answer: B

65. Which of the following is an example of a good prompt in prompt engineering?

A) "Tell me about AI."
 B) "Describe the applications of AI in healthcare and provide examples."
C) "What is AI?"
D) "Generate a sentence."
Answer: B

66. In prompt engineering, what is meant by "prompt tuning"?

A) Creating custom prompts to improve model performance
B) Tuning hyperparameters for model training
C) Fine-tuning the weights of a pre-trained model
D) Adjusting the training dataset
Answer: A

✅ Unit VI – ChatGPT Advanced Data Analysis and Advanced Topics in AI (Additional MCQs)

67. What is the primary advantage of using ChatGPT for data analysis?
A) It can generate real-time predictions
B) It can automate the process of data visualization and reporting
C) It can handle large datasets with ease
D) It works only with structured data
Answer: B

68. Which of the following is an important consideration when using AI in real-time

applications?
A) Model interpretability
B) Data privacy and security
C) Availability of labeled data
D) Model accuracy only
Answer: B

69. Which of the following is NOT typically a task in Natural Language Processing?
A) Sentiment analysis
B) Text summarization
C) Predictive maintenance
D) Named entity recognition
Answer: C

70. What is the goal of reinforcement learning in the context of game playing?
A) To optimize the agent’s strategy through rewards and penalties
B) To build a game-playing bot
C) To simulate human decision-making
D) To generate random strategies
Answer: A

71. Which of the following AI concepts is closely related to decision-making processes that are
probabilistic?
A) Bayesian networks
B) Neural networks
C) Genetic algorithms
 D) Min-Max algorithm
Answer: A

72. What is a potential drawback of using deep learning models?

A) They require a large amount of data for training
B) They cannot be applied to complex data types
C) They are not scalable
D) They always outperform traditional machine learning models
Answer: A