Artificial Intelligence MCQ’s

Artificial Intelligence MCQ's
Artificial Intelligence – 500 MCQs

Artificial Intelligence – 500 MCQs

Answers are hidden by default. Students can read the question first and then click to reveal the answer.

Showing 500 questions
Question 1
Introduction to Artificial Intelligence
Artificial Intelligence is a branch of:
  1. Physics
  2. Computer Science
  3. Chemistry
  4. Biology
Correct Answer: B) Computer Science
Question 2
Introduction to Artificial Intelligence
The term 'Artificial Intelligence' was first introduced at:
  1. MIT Conference 1960
  2. Dartmouth Conference 1956
  3. Stanford Conference 1970
  4. IBM Conference 1945
Correct Answer: B) Dartmouth Conference 1956
Question 3
Introduction to Artificial Intelligence
Alan Turing proposed the concept of machine intelligence in:
  1. 1945
  2. 1960
  3. 1950
  4. 1975
Correct Answer: C) 1950
Question 4
Introduction to Artificial Intelligence
AI systems differ from traditional programs because AI:
  1. Uses more RAM
  2. Can learn from data
  3. Has no bugs
  4. Uses faster CPUs
Correct Answer: B) Can learn from data
Question 5
Introduction to Artificial Intelligence
Which of the following is a goal of AI?
  1. Making faster printers
  2. Automating car engines
  3. Enabling machines to simulate human intelligence
  4. Designing keyboards
Correct Answer: C) Enabling machines to simulate human intelligence
Question 6
Introduction to Artificial Intelligence
Which is NOT an application of AI?
  1. Self-driving cars
  2. Chatbots
  3. Calculator
  4. Medical diagnosis systems
Correct Answer: C) Calculator
Question 7
Introduction to Artificial Intelligence
The Turing Test was proposed by:
  1. Marvin Minsky
  2. John McCarthy
  3. Alan Turing
  4. Geoffrey Hinton
Correct Answer: C) Alan Turing
Question 8
Introduction to Artificial Intelligence
The Turing Test evaluates a machine's ability to:
  1. Solve math problems
  2. Imitate human conversation
  3. Run faster programs
  4. Store large data
Correct Answer: B) Imitate human conversation
Question 9
Introduction to Artificial Intelligence
In traditional programming, the output is determined by:
  1. Data patterns
  2. Machine learning
  3. Predefined rules
  4. Neural networks
Correct Answer: C) Predefined rules
Question 10
Introduction to Artificial Intelligence
Which AI milestone involved defeating a world chess champion?
  1. AlphaGo
  2. IBM Deep Blue
  3. ELIZA
  4. Siri
Correct Answer: B) IBM Deep Blue
Question 11
Introduction to Artificial Intelligence
Which AI milestone involved defeating a world Go champion?
  1. IBM Deep Blue
  2. Google AlphaGo
  3. GPT
  4. Watson
Correct Answer: B) Google AlphaGo
Question 12
Introduction to Artificial Intelligence
AI systems can:
  1. Only follow fixed rules
  2. Never learn
  3. Analyze large data and identify patterns
  4. Only play games
Correct Answer: C) Analyze large data and identify patterns
Question 13
Introduction to Artificial Intelligence
The Dartmouth Conference is known for:
  1. Inventing the internet
  2. Officially introducing the term AI
  3. Building the first robot
  4. Creating Python language
Correct Answer: B) Officially introducing the term AI
Question 14
Introduction to Artificial Intelligence
Machine Learning became popular in which decade?
  1. 1950s
  2. 1960s
  3. 1980s
  4. 2010s
Correct Answer: C) 1980s
Question 15
Introduction to Artificial Intelligence
Deep Learning is a part of:
  1. Hardware engineering
  2. Networking
  3. Artificial Intelligence
  4. Database management
Correct Answer: C) Artificial Intelligence
Question 16
Introduction to Artificial Intelligence
Which is a key property of AI systems?
  1. Fixed output
  2. No data processing
  3. Ability to adapt to new inputs SECTION 2: Key Components of AI (Q21–Q35)
  4. Manual updates only
Correct Answer: C) Ability to adapt to new inputs SECTION 2: Key Components of AI (Q21–Q35)
Question 17
Key Components of AI
Spam detection using AI learns from:
  1. Manually written rules only
  2. Thousands of email examples
  3. Hardware settings
  4. Network configurations
Correct Answer: B) Thousands of email examples
Question 18
Key Components of AI
Computer Vision allows machines to:
  1. Hear sounds
  2. Process text
  3. Interpret images and videos
  4. Compile programs
Correct Answer: C) Interpret images and videos
Question 19
Key Components of AI
Natural Language Processing (NLP) deals with:
  1. Graphics rendering
  2. Human language understanding
  3. Hardware interfacing
  4. Memory management
Correct Answer: B) Human language understanding
Question 20
Key Components of AI
Robotics integrates AI with:
  1. Mathematics only
  2. Software compilers
  3. Mechanical/physical systems
  4. Database systems
Correct Answer: C) Mechanical/physical systems
Question 21
Key Components of AI
Which component of AI enables systems to learn without being explicitly programmed?
  1. Robotics
  2. Machine Learning
  3. Computer Vision
  4. Networking
Correct Answer: B) Machine Learning
Question 22
Key Components of AI
An intelligent agent perceives its environment through:
  1. Actuators
  2. Compilers
  3. Sensors
  4. Databases
Correct Answer: C) Sensors
Question 23
Key Components of AI
An intelligent agent takes actions through:
  1. Sensors
  2. Actuators
  3. Memory
  4. Algorithms
Correct Answer: B) Actuators
Question 24
Key Components of AI
Knowledge representation stores information as:
  1. Audio files
  2. Logical rules and knowledge bases
  3. Video streams
  4. Executable programs
Correct Answer: B) Logical rules and knowledge bases
Question 25
Key Components of AI
Which reasoning type derives specific conclusions from general rules?
  1. Inductive
  2. Random
  3. Deductive
  4. Statistical
Correct Answer: C) Deductive
Question 26
Key Components of AI
BFS stands for:
  1. Binary File System
  2. Breadth First Search
  3. Basic File Storage
  4. Boolean Function Set
Correct Answer: B) Breadth First Search
Question 27
Key Components of AI
DFS stands for:
  1. Data File Storage
  2. Depth First Search
  3. Distributed File System
  4. Dynamic Function Search
Correct Answer: B) Depth First Search
Question 28
Key Components of AI
A* is a type of:
  1. Uninformed search
  2. Random search
  3. Informed search algorithm
  4. Database query
Correct Answer: C) Informed search algorithm
Question 29
Key Components of AI
Facial recognition is an application of:
  1. NLP
  2. Robotics
  3. Computer Vision
  4. Machine Learning only
Correct Answer: C) Computer Vision
Question 30
Key Components of AI
Which component bridges communication between humans and machines?
  1. Computer Vision
  2. Robotics
  3. NLP
  4. Search algorithms
Correct Answer: C) NLP
Question 31
Key Components of AI
Types of Machine Learning include:
  1. Supervised, Unsupervised, Reinforcement
  2. Basic, Advanced, Expert
  3. Online, Offline, Hybrid
  4. Fast, Slow, Medium
Correct Answer: A) Supervised, Unsupervised, Reinforcement
Question 32
Key Components of AI
Medical image diagnosis uses:
  1. NLP
  2. Computer Vision
  3. Robotics
  4. Knowledge representation only
Correct Answer: B) Computer Vision
Question 33
Key Components of AI
Semantic networks are used in:
  1. Networking hardware SECTION 3: Intelligent Agents (Q36–Q55)
  2. Knowledge representation
  3. Compiler design
  4. Operating systems
Correct Answer: B) Knowledge representation
Question 34
Intelligent Agents
Probabilistic reasoning is used when:
  1. Certainty is guaranteed
  2. Dealing with uncertainty
  3. Hardware is unreliable
  4. Programs crash
Correct Answer: B) Dealing with uncertainty
Question 35
Intelligent Agents
Which search algorithm uses f(n) = g(n) + h(n)?
  1. BFS
  2. DFS
  3. A*
  4. DLS
Correct Answer: C) A*
Question 36
Intelligent Agents
An intelligent agent is:
  1. A fixed program
  2. An autonomous system that perceives and acts
  3. A database
  4. A compiler
Correct Answer: B) An autonomous system that perceives and acts
Question 37
Intelligent Agents
The basic cycle of an intelligent agent is:
  1. Input-Output
  2. Compile-Run-Debug
  3. Perception-Decision-Action
  4. Read-Write-Execute
Correct Answer: C) Perception-Decision-Action
Question 38
Intelligent Agents
Simple reflex agents act based on:
  1. Past memory
  2. Future planning
  3. Current percept only
  4. Random decisions
Correct Answer: C) Current percept only
Question 39
Intelligent Agents
Model-based agents differ from simple reflex agents because they:
  1. Never use sensors
  2. Maintain an internal state/model
  3. Are slower
  4. Use less memory
Correct Answer: B) Maintain an internal state/model
Question 40
Intelligent Agents
Goal-based agents use:
  1. Random actions
  2. Only current input
  3. Planning to achieve goals
  4. Fixed rules only
Correct Answer: C) Planning to achieve goals
Question 41
Intelligent Agents
A utility-based agent aims to:
  1. Minimize utility
  2. Maximize utility function
  3. Ignore trade-offs
  4. Always achieve one goal
Correct Answer: B) Maximize utility function
Question 42
Intelligent Agents
Learning agents improve by:
  1. Increasing hardware
  2. Learning from experience and feedback
  3. Removing old rules
  4. Random changes
Correct Answer: B) Learning from experience and feedback
Question 43
Intelligent Agents
A thermostat is an example of:
  1. Goal-based agent
  2. Learning agent
  3. Simple reflex agent
  4. Utility-based agent
Correct Answer: C) Simple reflex agent
Question 44
Intelligent Agents
Google Maps route planning is an example of:
  1. Simple reflex agent
  2. Goal-based agent
  3. Model-based agent
  4. Utility-based agent
Correct Answer: B) Goal-based agent
Question 45
Intelligent Agents
Netflix recommendation system is an example of:
  1. Simple reflex agent
  2. Goal-based agent
  3. Learning agent
  4. Model-based agent
Correct Answer: C) Learning agent
Question 46
Intelligent Agents
A rational agent selects actions that:
  1. Always guarantee success
  2. Maximize expected performance
  3. Are cheapest
  4. Are random
Correct Answer: B) Maximize expected performance
Question 47
Intelligent Agents
Rationality does NOT mean:
  1. Best possible decision
  2. Using available information
  3. Always being perfect
  4. Maximizing performance measure
Correct Answer: C) Always being perfect
Question 48
Intelligent Agents
PEAS stands for:
  1. Process, Execute, Analyze, System
  2. Performance, Environment, Actuators, Sensors
  3. Program, Engine, Agent, Storage
  4. Perception, Execution, Action, State
Correct Answer: B) Performance, Environment, Actuators, Sensors
Question 49
Intelligent Agents
In PEAS for a self-driving car, Actuators include:
  1. Cameras
  2. GPS
  3. Steering and brakes
  4. Radar
Correct Answer: C) Steering and brakes
Question 50
Knowledge-Based Systems & ELIZA
Performance measure for a spam filter is:
  1. Speed of deletion
  2. High detection accuracy, low false positives
  3. Email storage size
  4. Number of emails received
Correct Answer: B) High detection accuracy, low false positives
Question 51
Knowledge-Based Systems & ELIZA
Sensors of a self-driving car include:
  1. Steering wheel and brakes
  2. Cameras, radar, GPS, LiDAR
  3. Engine and fuel
  4. Seats and windows
Correct Answer: B) Cameras, radar, GPS, LiDAR
Question 52
Knowledge-Based Systems & ELIZA
Which agent type is best for uncertain environments with trade-offs?
  1. Simple reflex agent
  2. Model-based agent
  3. Utility-based agent
  4. Compiler agent
Correct Answer: C) Utility-based agent
Question 53
Knowledge-Based Systems & ELIZA
The environment of a spam filter is:
  1. Roads and traffic
  2. Incoming emails
  3. Medical records
  4. Stock market data
Correct Answer: B) Incoming emails
Question 54
Knowledge-Based Systems & ELIZA
Which is NOT a type of intelligent agent?
  1. Simple reflex agent
  2. Goal-based agent
  3. Compiler-based agent
  4. Learning agent
Correct Answer: C) Compiler-based agent
Question 55
Knowledge-Based Systems & ELIZA
Actuators help an intelligent agent to:
  1. Collect input data
  2. Process algorithms
  3. Execute actions in environment
  4. Store memory
Correct Answer: C) Execute actions in environment
Question 56
Knowledge-Based Systems & ELIZA
A Knowledge-Based System uses:
  1. Only databases
  2. Stored knowledge and logical reasoning
  3. Random decision making
  4. Hardware components only
Correct Answer: B) Stored knowledge and logical reasoning
Question 57
Knowledge-Based Systems & ELIZA
The inference engine in a KBS is responsible for:
  1. Collecting data from sensors
  2. Applying logic to derive conclusions
  3. Storing facts and rules
  4. Printing results
Correct Answer: B) Applying logic to derive conclusions
Question 58
Knowledge-Based Systems & ELIZA
Forward chaining starts from:
  1. Goal
  2. Random point
  3. Known facts
  4. Conclusion
Correct Answer: C) Known facts
Question 59
Knowledge-Based Systems & ELIZA
Backward chaining starts from:
  1. Known facts
  2. Goal
  3. Random variables
  4. Sensors
Correct Answer: B) Goal
Question 60
Knowledge-Based Systems & ELIZA
ELIZA was developed by:
  1. Alan Turing
  2. John McCarthy
  3. Joseph Weizenbaum
  4. Marvin Minsky
Correct Answer: C) Joseph Weizenbaum
Question 61
Knowledge-Based Systems & ELIZA
ELIZA was developed in the:
  1. 1950s
  2. 1960s
  3. 1980s
  4. 2000s
Correct Answer: B) 1960s
Question 62
Knowledge-Based Systems & ELIZA
ELIZA simulated a:
  1. Teacher
  2. Doctor
  3. Psychotherapist
  4. Lawyer
Correct Answer: C) Psychotherapist
Question 63
Knowledge-Based Systems & ELIZA
ELIZA used which technique?
  1. Deep learning
  2. Neural networks
  3. Pattern matching
  4. Reinforcement learning
Correct Answer: C) Pattern matching
Question 64
Knowledge-Based Systems & ELIZA
ELIZA demonstrated:
  1. True language understanding
  2. That AI can simulate conversation
  3. Deep learning capabilities
  4. Robotic movement
Correct Answer: B) That AI can simulate conversation
Question 65
Knowledge-Based Systems & ELIZA
The knowledge base in a KBS stores:
  1. Programs and executables
  2. Facts and rules
  3. Hardware data
  4. Image files
Correct Answer: B) Facts and rules
Question 66
Knowledge-Based Systems & ELIZA
An expert system is a type of:
  1. Game
  2. Knowledge-Based System
  3. Database
  4. Compiler SECTION 5: Knowledge Representation & Logic (Q71–Q95)
Correct Answer: B) Knowledge-Based System
Question 67
Knowledge Representation & Logic
Working memory in a KBS stores:
  1. Long-term facts
  2. Current input data for processing
  3. Historical data
  4. Hardware specifications
Correct Answer: B) Current input data for processing
Question 68
Knowledge Representation & Logic
IF fever AND cough implies possible flu is an example of:
  1. A search algorithm
  2. A knowledge base rule
  3. A sorting method
  4. A database query
Correct Answer: B) A knowledge base rule
Question 69
Knowledge Representation & Logic
KBS systems were popular mainly in:
  1. 1950s-1960s
  2. 1970s-1980s
  3. 1990s-2000s
  4. 2010s-2020s
Correct Answer: B) 1970s-1980s
Question 70
Knowledge Representation & Logic
The explanation system in KBS is used to:
  1. Store rules
  2. Explain why a conclusion was reached
  3. Process sensor data
  4. Compress data
Correct Answer: B) Explain why a conclusion was reached
Question 71
Knowledge Representation & Logic
Knowledge Representation in AI is used to:
  1. Increase memory size
  2. Store structured information for reasoning
  3. Speed up hardware
  4. Design interfaces
Correct Answer: B) Store structured information for reasoning
Question 72
Knowledge Representation & Logic
A proposition is a statement that is:
  1. Always true
  2. Always false
  3. Either true or false
  4. Partially true
Correct Answer: C) Either true or false
Question 73
Knowledge Representation & Logic
Propositional Logic deals with:
  1. Complex relationships
  2. Simple true/false statements
  3. Variable relationships
  4. Probabilistic statements
Correct Answer: B) Simple true/false statements
Question 74
Knowledge Representation & Logic
The symbol AND (conjunction) represents:
  1. OR
  2. NOT
  3. AND
  4. IMPLIES
Correct Answer: C) AND
Question 75
Knowledge Representation & Logic
The symbol OR (disjunction) represents:
  1. AND
  2. OR
  3. NOT
  4. IF-THEN
Correct Answer: B) OR
Question 76
Knowledge Representation & Logic
The symbol NOT (negation) represents:
  1. AND
  2. OR
  3. NOT
  4. IMPLIES
Correct Answer: C) NOT
Question 77
Knowledge Representation & Logic
P IMPLIES Q means:
  1. P and Q
  2. P or Q
  3. If P then Q
  4. Not P
Correct Answer: C) If P then Q
Question 78
Knowledge Representation & Logic
Predicate Logic is also called:
  1. Boolean Logic
  2. First-Order Logic
  3. Digital Logic
  4. Binary Logic
Correct Answer: B) First-Order Logic
Question 79
Knowledge Representation & Logic
The universal quantifier (for all) means:
  1. There exists
  2. For all
  3. If-then
  4. And
Correct Answer: B) For all
Question 80
Knowledge Representation & Logic
The existential quantifier (there exists) means:
  1. For all
  2. If-then
  3. There exists
  4. And
Correct Answer: C) There exists
Question 81
Knowledge Representation & Logic
ForAll x (Human(x) implies Mortal(x)) is an example of:
  1. Propositional Logic
  2. Predicate Logic
  3. Fuzzy Logic
  4. Binary Logic
Correct Answer: B) Predicate Logic
Question 82
Knowledge Representation & Logic
Which logic cannot represent relationships between objects?
  1. Predicate Logic
  2. First-Order Logic
  3. Propositional Logic
  4. Modal Logic
Correct Answer: C) Propositional Logic
Question 83
Knowledge Representation & Logic
Predicate Logic is more expressive because it allows:
  1. Only true/false statements
  2. Objects, variables, relationships, quantifiers
  3. Only mathematical expressions
  4. Only numeric data
Correct Answer: B) Objects, variables, relationships, quantifiers
Question 84
Knowledge Representation & Logic
Modus Ponens is an example of:
  1. Search algorithm
  2. Inference rule
  3. Data structure
  4. Sorting method
Correct Answer: B) Inference rule
Question 85
Knowledge Representation & Logic
If P implies Q and P is true, then Q is:
  1. False
  2. Unknown
  3. True
  4. Undefined
Correct Answer: C) True
Question 86
Knowledge Representation & Logic
Deductive reasoning guarantees:
  1. Probable conclusions
  2. Logically certain conclusions
  3. Approximate results
  4. Random outputs
Correct Answer: B) Logically certain conclusions
Question 87
Knowledge Representation & Logic
Inductive reasoning is:
  1. Always certain
  2. Probabilistic
  3. Guaranteed correct
  4. Mathematical only
Correct Answer: B) Probabilistic
Question 88
Knowledge Representation & Logic
Which reasoning moves from specific to general?
  1. Deductive
  2. Forward chaining
  3. Inductive
  4. Backward chaining
Correct Answer: C) Inductive
Question 89
Knowledge Representation & Logic
Forward chaining is also called:
  1. Goal-driven
  2. Data-driven
  3. Random-driven
  4. Rule-driven
Correct Answer: B) Data-driven
Question 90
Knowledge Representation & Logic
Backward chaining is commonly used in:
  1. Expert systems only
  2. Prolog and medical diagnosis
  3. Networking
  4. Sorting algorithms
Correct Answer: B) Prolog and medical diagnosis
Question 91
Knowledge Representation & Logic
Inference mechanisms help AI systems to:
  1. Store data
  2. Derive new knowledge
  3. Increase RAM
  4. Compress files
Correct Answer: B) Derive new knowledge
Question 92
Knowledge Representation & Logic
Constants in Predicate Logic represent: SECTION 6: Uninformed Search — BFS, DFS, DLS, IDS (Q96–Q130)
  1. Variables
  2. Specific objects like Ali or Car1
  3. Operations
  4. Rules
Correct Answer: B) Specific objects like Ali or Car1
Question 93
Uninformed Search — BFS, DFS, DLS, IDS
Which reasoning is used mainly in Machine Learning?
  1. Deductive
  2. Inductive
  3. Symbolic
  4. Boolean
Correct Answer: B) Inductive
Question 94
Uninformed Search — BFS, DFS, DLS, IDS
Inferential adequacy means:
  1. Fast hardware
  2. Ability to derive new knowledge
  3. Large storage
  4. Network access
Correct Answer: B) Ability to derive new knowledge
Question 95
Uninformed Search — BFS, DFS, DLS, IDS
Ontologies in AI are used for:
  1. Sorting data
  2. Representing structured domain knowledge
  3. Networking
  4. Encryption
Correct Answer: B) Representing structured domain knowledge
Question 96
Uninformed Search — BFS, DFS, DLS, IDS
Uninformed search is also called:
  1. Guided search
  2. Blind search
  3. Intelligent search
  4. Optimal search
Correct Answer: B) Blind search
Question 97
Uninformed Search — BFS, DFS, DLS, IDS
BFS explores nodes:
  1. Deepest first
  2. Randomly
  3. Level by level
  4. In reverse order
Correct Answer: C) Level by level
Question 98
Uninformed Search — BFS, DFS, DLS, IDS
BFS uses which data structure?
  1. Stack
  2. Queue
  3. Tree
  4. Array
Correct Answer: B) Queue
Question 99
Uninformed Search — BFS, DFS, DLS, IDS
DFS uses which data structure?
  1. Queue
  2. Stack
  3. Heap
  4. Linked list
Correct Answer: B) Stack
Question 100
Uninformed Search — BFS, DFS, DLS, IDS
Queue follows:
  1. LIFO
  2. FIFO
  3. Random order
  4. Priority order
Correct Answer: B) FIFO
Question 101
Uninformed Search — BFS, DFS, DLS, IDS
Stack follows:
  1. FIFO
  2. LIFO
  3. Random order
  4. Alphabetical order
Correct Answer: B) LIFO
Question 102
Uninformed Search — BFS, DFS, DLS, IDS
BFS is complete meaning:
  1. It always finds goal if it exists
  2. It is always optimal
  3. It uses less memory
  4. It explores all depths
Correct Answer: A) It always finds goal if it exists
Question 103
Uninformed Search — BFS, DFS, DLS, IDS
BFS is optimal when:
  1. All step costs are equal
  2. Tree is balanced
  3. Memory is unlimited
  4. Goal is at root
Correct Answer: A) All step costs are equal
Question 104
Uninformed Search — BFS, DFS, DLS, IDS
Time complexity of BFS is:
  1. O(n)
  2. O(log n)
  3. O(b^d)
  4. O(b)
Correct Answer: C) O(b^d)
Question 105
Uninformed Search — BFS, DFS, DLS, IDS
Space complexity of BFS is:
  1. O(bd)
  2. O(b^d)
  3. O(bm)
  4. O(b^L)
Correct Answer: B) O(b^d)
Question 106
Uninformed Search — BFS, DFS, DLS, IDS
In BFS, 'b' represents:
  1. Depth of goal
  2. Branching factor
  3. Number of edges
  4. Bandwidth
Correct Answer: B) Branching factor
Question 107
Uninformed Search — BFS, DFS, DLS, IDS
In BFS, 'd' represents:
  1. Branching factor
  2. Data size
  3. Depth of goal node
  4. Distance traveled
Correct Answer: C) Depth of goal node
Question 108
Uninformed Search — BFS, DFS, DLS, IDS
DFS explores:
  1. Level by level
  2. Randomly
  3. Deepest nodes first
  4. Shallowest first
Correct Answer: C) Deepest nodes first
Question 109
Uninformed Search — BFS, DFS, DLS, IDS
DFS is complete?
  1. Yes always
  2. Yes if depth is limited
  3. No
  4. Yes if tree is small
Correct Answer: C) No
Question 110
Uninformed Search — BFS, DFS, DLS, IDS
DFS is optimal?
  1. Yes always
  2. Only with equal costs
  3. No
  4. Yes with heuristics
Correct Answer: C) No
Question 111
Uninformed Search — BFS, DFS, DLS, IDS
Time complexity of DFS is:
  1. O(b^d)
  2. O(b^m)
  3. O(n)
  4. O(log n)
Correct Answer: B) O(b^m)
Question 112
Uninformed Search — BFS, DFS, DLS, IDS
In DFS, 'm' represents:
  1. Memory size
  2. Maximum depth
  3. Minimum depth
  4. Mode factor
Correct Answer: B) Maximum depth
Question 113
Uninformed Search — BFS, DFS, DLS, IDS
DFS may get stuck in:
  1. Shortest paths
  2. Goal states
  3. Infinite paths
  4. Optimal solutions
Correct Answer: C) Infinite paths
Question 114
Uninformed Search — BFS, DFS, DLS, IDS
DLS stands for:
  1. Deep Learning Search
  2. Depth Limited Search
  3. Data Level Search
  4. Dynamic List Sort
Correct Answer: B) Depth Limited Search
Question 115
Uninformed Search — BFS, DFS, DLS, IDS
DLS is a modified version of:
  1. BFS
  2. DFS
  3. IDS
  4. A*
Correct Answer: B) DFS
Question 116
Uninformed Search — BFS, DFS, DLS, IDS
DLS prevents:
  1. Finding solutions
  2. Infinite search paths
  3. Memory usage
  4. Level-order traversal
Correct Answer: B) Infinite search paths
Question 117
Uninformed Search — BFS, DFS, DLS, IDS
IDS stands for:
  1. Integrated Data System
  2. Iterative Deepening Search
  3. Indexed Data Store
  4. Inverse Depth Search
Correct Answer: B) Iterative Deepening Search
Question 118
Uninformed Search — BFS, DFS, DLS, IDS
IDS combines:
  1. BFS and DFS
  2. BFS and A*
  3. DFS and heuristics
  4. A* and greedy
Correct Answer: A) BFS and DFS
Question 119
Uninformed Search — BFS, DFS, DLS, IDS
IDS gradually increases:
  1. Memory size
  2. Branching factor
  3. Depth limit
  4. Node count
Correct Answer: C) Depth limit
Question 120
Uninformed Search — BFS, DFS, DLS, IDS
IDS is complete?
  1. No
  2. Only sometimes
  3. Yes
  4. Only with heuristics
Correct Answer: C) Yes
Question 121
Uninformed Search — BFS, DFS, DLS, IDS
IDS is optimal when:
  1. Memory is unlimited
  2. Step costs are equal
  3. Tree is balanced
  4. Depth is 0
Correct Answer: B) Step costs are equal
Question 122
Uninformed Search — BFS, DFS, DLS, IDS
Space complexity of IDS is:
  1. O(b^d)
  2. O(bm)
  3. O(bd)
  4. O(b^L)
Correct Answer: C) O(bd)
Question 123
Uninformed Search — BFS, DFS, DLS, IDS
Which algorithm uses least memory?
  1. BFS
  2. DFS
  3. IDS
  4. A*
Correct Answer: B) DFS
Question 124
Uninformed Search — BFS, DFS, DLS, IDS
A state space represents:
  1. RAM capacity
  2. All possible states of a problem
  3. Network topology
  4. Database schema
Correct Answer: B) All possible states of a problem
Question 125
Uninformed Search — BFS, DFS, DLS, IDS
The initial state in a search problem is:
  1. The goal
  2. The starting configuration
  3. A random state
  4. The last state
Correct Answer: B) The starting configuration
Question 126
Uninformed Search — BFS, DFS, DLS, IDS
Goal test in a search problem: SECTION 7: Informed Search — Heuristics, Greedy, A* (Q131–Q155)
  1. Generates new states
  2. Checks if goal is reached
  3. Sorts states
  4. Deletes states
Correct Answer: B) Checks if goal is reached
Question 127
Informed Search — Heuristics, Greedy, A*
Path cost in search represents:
  1. Memory used
  2. Time taken
  3. Cost to reach goal state
  4. Number of nodes
Correct Answer: C) Cost to reach goal state
Question 128
Informed Search — Heuristics, Greedy, A*
Which algorithm guarantees shortest path with equal costs?
  1. DFS
  2. DLS
  3. BFS
  4. IDS
Correct Answer: C) BFS
Question 129
Informed Search — Heuristics, Greedy, A*
Backtracking in DFS means:
  1. Going to root
  2. Going to a previous state when stuck
  3. Restarting search
  4. Deleting the tree
Correct Answer: B) Going to a previous state when stuck
Question 130
Informed Search — Heuristics, Greedy, A*
State space is usually represented as:
  1. A spreadsheet
  2. A tree or graph
  3. A table
  4. An array
Correct Answer: B) A tree or graph
Question 131
Informed Search — Heuristics, Greedy, A*
Informed search is also known as:
  1. Blind search
  2. Heuristic search
  3. Random search
  4. Exhaustive search
Correct Answer: B) Heuristic search
Question 132
Informed Search — Heuristics, Greedy, A*
A heuristic function is represented as:
  1. g(n)
  2. f(n)
  3. h(n)
  4. c(n)
Correct Answer: C) h(n)
Question 133
Informed Search — Heuristics, Greedy, A*
h(n) estimates:
  1. Distance from start to node
  2. Cost of entire graph
  3. Distance from node to goal
  4. Memory usage
Correct Answer: C) Distance from node to goal
Question 134
Informed Search — Heuristics, Greedy, A*
Greedy Best First Search selects:
  1. Node with highest h(n)
  2. Node with lowest h(n)
  3. Random node
  4. Deepest node
Correct Answer: B) Node with lowest h(n)
Question 135
Informed Search — Heuristics, Greedy, A*
Evaluation function of Greedy Best First Search is:
  1. f(n) = g(n) + h(n)
  2. f(n) = g(n)
  3. f(n) = h(n)
  4. f(n) = h(n) – g(n)
Correct Answer: C) f(n) = h(n)
Question 136
Informed Search — Heuristics, Greedy, A*
A* algorithm evaluation function is:
  1. f(n) = g(n)
  2. f(n) = h(n)
  3. f(n) = g(n) + h(n)
  4. f(n) = g(n) * h(n)
Correct Answer: C) f(n) = g(n) + h(n)
Question 137
Informed Search — Heuristics, Greedy, A*
In A*, g(n) represents:
  1. Estimated cost to goal
  2. Cost from start to current node
  3. Total nodes visited
  4. Heuristic value
Correct Answer: B) Cost from start to current node
Question 138
Informed Search — Heuristics, Greedy, A*
In A*, h(n) represents:
  1. Actual path cost
  2. Total cost
  3. Heuristic estimate to goal
  4. Node weight
Correct Answer: C) Heuristic estimate to goal
Question 139
Informed Search — Heuristics, Greedy, A*
A* is optimal when the heuristic is:
  1. Consistent only
  2. Admissible
  3. Always zero
  4. Always maximum
Correct Answer: B) Admissible
Question 140
Informed Search — Heuristics, Greedy, A*
An admissible heuristic:
  1. Always overestimates
  2. Never overestimates the true cost
  3. Is always zero
  4. Is always maximum
Correct Answer: B) Never overestimates the true cost
Question 141
Informed Search — Heuristics, Greedy, A*
Formally, admissible means h(n):
  1. >= h*(n)
  2. <= h*(n)
  3. = 0
  4. = g(n)
Correct Answer: B) <= h*(n)
Question 142
Informed Search — Heuristics, Greedy, A*
Manhattan Distance is used in:
  1. Road navigation
  2. Grid-based problems like 8-puzzle
  3. 3D robotics
  4. Text processing
Correct Answer: B) Grid-based problems like 8-puzzle
Question 143
Informed Search — Heuristics, Greedy, A*
Manhattan Distance formula is:
  1. sqrt sum of squares
  2. |x1-x2| + |y1-y2|
  3. (x1-x2) * (y1-y2)
  4. x1*x2 + y1*y2
Correct Answer: B) |x1-x2| + |y1-y2|
Question 144
Informed Search — Heuristics, Greedy, A*
Euclidean Distance is used in:
  1. Grid problems only
  2. Text search
  3. Continuous movement like robotics
  4. Graph databases
Correct Answer: C) Continuous movement like robotics
Question 145
Informed Search — Heuristics, Greedy, A*
Straight-line distance is admissible because:
  1. Roads are always straight
  2. Actual road distance cannot be less than straight-line
  3. It always overestimates
  4. It equals actual distance
Correct Answer: B) Actual road distance cannot be less than straight-line
Question 146
Informed Search — Heuristics, Greedy, A*
A* algorithm is used in:
  1. Text editing
  2. File compression
  3. GPS navigation and pathfinding
  4. Email management
Correct Answer: C) GPS navigation and pathfinding
Question 147
Informed Search — Heuristics, Greedy, A*
Greedy search may get stuck in:
  1. Optimal solutions
  2. Local minima
  3. Shortest paths
  4. Goal states
Correct Answer: B) Local minima
Question 148
Informed Search — Heuristics, Greedy, A*
A consistent heuristic satisfies:
  1. h(n) = 0 always
  2. h(n) <= c(n,n') + h(n')
  3. h(n) >= h*(n)
  4. h(n) = g(n)
Correct Answer: B) h(n) <= c(n,n') + h(n')
Question 149
Informed Search — Heuristics, Greedy, A*
All consistent heuristics are:
  1. Not admissible
  2. Admissible
  3. Always zero
  4. Always maximum
Correct Answer: B) Admissible
Question 150
Informed Search — Heuristics, Greedy, A*
A* uses a priority queue ordered by:
  1. g(n)
  2. h(n)
  3. f(n) = g(n) + h(n)
  4. Random order
Correct Answer: C) f(n) = g(n) + h(n)
Question 151
Informed Search — Heuristics, Greedy, A*
Which algorithm is NOT an informed search?
  1. A*
  2. Greedy Best First Search
  3. BFS SECTION 8: Local Search — Hill Climbing, SA, Genetic Algorithms (Q156–Q180)
  4. Heuristic search
Correct Answer: C) BFS SECTION 8: Local Search — Hill Climbing, SA, Genetic Algorithms (Q156–Q180)
Question 152
Local Search — Hill Climbing, SA, Genetic Algorithms
Greedy Best First Search is:
  1. Optimal
  2. Complete always
  3. Fast but not always optimal
  4. Always finds shortest path
Correct Answer: C) Fast but not always optimal
Question 153
Local Search — Hill Climbing, SA, Genetic Algorithms
A* combines advantages of:
  1. DFS and BFS
  2. Uniform Cost Search and Greedy Search
  3. DLS and IDS
  4. BFS and DLS
Correct Answer: B) Uniform Cost Search and Greedy Search
Question 154
Local Search — Hill Climbing, SA, Genetic Algorithms
Video game pathfinding uses:
  1. Only BFS
  2. Only DFS
  3. Informed search like A*
  4. Random search
Correct Answer: C) Informed search like A*
Question 155
Local Search — Hill Climbing, SA, Genetic Algorithms
A good heuristic should:
  1. Always overestimate
  2. Be hard to compute
  3. Provide accurate estimates and easy to compute
  4. Always return zero
Correct Answer: C) Provide accurate estimates and easy to compute
Question 156
Local Search — Hill Climbing, SA, Genetic Algorithms
Local search focuses on:
  1. Exploring all paths
  2. Improving a single current solution
  3. Level-by-level expansion
  4. Random exploration
Correct Answer: B) Improving a single current solution
Question 157
Local Search — Hill Climbing, SA, Genetic Algorithms
Hill Climbing is a:
  1. Random algorithm
  2. Greedy local search algorithm
  3. Complete search algorithm
  4. BFS variant
Correct Answer: B) Greedy local search algorithm
Question 158
Local Search — Hill Climbing, SA, Genetic Algorithms
Hill Climbing selects:
  1. Worst neighbor
  2. Random neighbor
  3. Best neighboring state
  4. All neighbors
Correct Answer: C) Best neighboring state
Question 159
Local Search — Hill Climbing, SA, Genetic Algorithms
Main drawback of Hill Climbing:
  1. High memory usage
  2. Slow speed
  3. Gets stuck in local maxima
  4. Complex implementation
Correct Answer: C) Gets stuck in local maxima
Question 160
Local Search — Hill Climbing, SA, Genetic Algorithms
A local maximum is:
  1. The global best solution
  2. A solution better than neighbors but not globally optimal
  3. A random peak
  4. The starting point
Correct Answer: B) A solution better than neighbors but not globally optimal
Question 161
Local Search — Hill Climbing, SA, Genetic Algorithms
A plateau in Hill Climbing is where:
  1. All neighbors are better
  2. No neighbors exist
  3. All neighbors have equal value
  4. Goal is found
Correct Answer: C) All neighbors have equal value
Question 162
Local Search — Hill Climbing, SA, Genetic Algorithms
Simulated Annealing is inspired by:
  1. Biology
  2. Metallurgy/Physics
  3. Chemistry
  4. Mathematics
Correct Answer: B) Metallurgy/Physics
Question 163
Local Search — Hill Climbing, SA, Genetic Algorithms
Simulated Annealing can accept:
  1. Only better solutions
  2. No solutions
  3. Worse solutions sometimes
  4. Only optimal solutions
Correct Answer: C) Worse solutions sometimes
Question 164
Local Search — Hill Climbing, SA, Genetic Algorithms
Temperature in Simulated Annealing controls:
  1. CPU speed
  2. Probability of accepting worse solutions
  3. Number of nodes
  4. Path cost
Correct Answer: B) Probability of accepting worse solutions
Question 165
Local Search — Hill Climbing, SA, Genetic Algorithms
As temperature decreases in Simulated Annealing:
  1. More bad moves accepted
  2. Algorithm stops
  3. Probability of accepting bad moves decreases
  4. Memory increases
Correct Answer: C) Probability of accepting bad moves decreases
Question 166
Local Search — Hill Climbing, SA, Genetic Algorithms
Simulated Annealing helps to:
  1. Increase speed always
  2. Escape local maxima
  3. Find exact solutions always
  4. Reduce nodes always
Correct Answer: B) Escape local maxima
Question 167
Local Search — Hill Climbing, SA, Genetic Algorithms
Genetic Algorithms are based on:
  1. Physics laws
  2. Natural selection and evolution
  3. Graph theory
  4. Sorting principles
Correct Answer: B) Natural selection and evolution
Question 168
Local Search — Hill Climbing, SA, Genetic Algorithms
In Genetic Algorithm, a solution is called:
  1. Node
  2. Chromosome
  3. Edge
  4. State
Correct Answer: B) Chromosome
Question 169
Local Search — Hill Climbing, SA, Genetic Algorithms
In Genetic Algorithm, fitness function:
  1. Generates nodes
  2. Evaluates solution quality
  3. Sorts chromosomes
  4. Stores values
Correct Answer: B) Evaluates solution quality
Question 170
Local Search — Hill Climbing, SA, Genetic Algorithms
Crossover in Genetic Algorithm means:
  1. Deleting a solution
  2. Copying a solution
  3. Combining two parent solutions
  4. Sorting solutions
Correct Answer: C) Combining two parent solutions
Question 171
Local Search — Hill Climbing, SA, Genetic Algorithms
Mutation in Genetic Algorithm means:
  1. Deleting solution
  2. Copying solution
  3. Random change in solution
  4. Sorting solution
Correct Answer: C) Random change in solution
Question 172
Local Search — Hill Climbing, SA, Genetic Algorithms
Genetic Algorithms work with:
  1. Single solution
  2. Population of solutions
  3. Graphs only
  4. Trees only
Correct Answer: B) Population of solutions
Question 173
Local Search — Hill Climbing, SA, Genetic Algorithms
Which algorithm can escape local maxima best?
  1. Hill Climbing
  2. BFS
  3. Simulated Annealing
  4. DFS
Correct Answer: C) Simulated Annealing
Question 174
Local Search — Hill Climbing, SA, Genetic Algorithms
Local search is best for:
  1. Pathfinding with exact paths
  2. Optimization problems
  3. Sorting data
  4. Searching arrays
Correct Answer: B) Optimization problems
Question 175
Local Search — Hill Climbing, SA, Genetic Algorithms
Hill Climbing stops when:
  1. Goal is found always
  2. No better neighbor exists
  3. Memory is full
  4. Time limit ends
Correct Answer: B) No better neighbor exists
Question 176
Local Search — Hill Climbing, SA, Genetic Algorithms
Which local search uses a population of solutions?
  1. Hill Climbing
  2. Simulated Annealing
  3. Genetic Algorithm
  4. DFS
Correct Answer: C) Genetic Algorithm
Question 177
AI Applications, Ethics & Society
Local search uses:
  1. Large amounts of memory
  2. Very little memory
  3. External databases
  4. Network resources
Correct Answer: B) Very little memory
Question 178
AI Applications, Ethics & Society
Ridge problem in Hill Climbing means:
  1. Algorithm finds global maximum
  2. Path to optimal solution is not straightforward
  3. No neighbors exist
  4. All neighbors are equal
Correct Answer: B) Path to optimal solution is not straightforward
Question 179
AI Applications, Ethics & Society
Genetic Algorithm is slower than Hill Climbing but:
  1. Uses more memory
  2. Finds worse solutions
  3. Is better at global optimization
  4. Is simpler
Correct Answer: C) Is better at global optimization
Question 180
AI Applications, Ethics & Society
Selection in Genetic Algorithm prefers:
  1. Random chromosomes
  2. Chromosomes with higher fitness
  3. Chromosomes with lower fitness
  4. Oldest chromosomes
Correct Answer: B) Chromosomes with higher fitness
Question 181
AI Applications, Ethics & Society
AI in healthcare is used for:
  1. Writing code
  2. Disease detection and medical imaging
  3. Playing games
  4. Making websites
Correct Answer: B) Disease detection and medical imaging
Question 182
AI Applications, Ethics & Society
AI in finance is used for:
  1. Image recognition
  2. Fraud detection and risk assessment
  3. Speech synthesis
  4. Video games
Correct Answer: B) Fraud detection and risk assessment
Question 183
AI Applications, Ethics & Society
Adaptive learning platforms use AI in:
  1. Transportation
  2. Healthcare
  3. Education
  4. Manufacturing
Correct Answer: C) Education
Question 184
AI Applications, Ethics & Society
Traffic management systems are AI in:
  1. Finance
  2. Education
  3. Transportation
  4. Healthcare
Correct Answer: C) Transportation
Question 185
AI Applications, Ethics & Society
Recommendation systems like Netflix use:
  1. Manual curation
  2. AI and Machine Learning
  3. Random selection
  4. User voting only
Correct Answer: B) AI and Machine Learning
Question 186
AI Applications, Ethics & Society
Self-driving cars use AI mainly for:
  1. Playing music
  2. Environmental perception and decision-making
  3. Typing documents
  4. Printing reports
Correct Answer: B) Environmental perception and decision-making
Question 187
AI Applications, Ethics & Society
Algorithmic trading uses AI in:
  1. Education sector
  2. Finance sector
  3. Healthcare sector
  4. Entertainment sector
Correct Answer: B) Finance sector
Question 188
AI Applications, Ethics & Society
Robotic surgery uses AI in:
  1. Finance
  2. Transportation
  3. Healthcare
  4. Education
Correct Answer: C) Healthcare
Question 189
AI Applications, Ethics & Society
Sentiment analysis is an application of:
  1. Computer Vision
  2. Robotics
  3. NLP
  4. Search algorithms
Correct Answer: C) NLP
Question 190
AI Applications, Ethics & Society
Machine translation is an application of:
  1. Computer Vision
  2. NLP
  3. Robotics
  4. Search algorithms
Correct Answer: B) NLP
Question 191
AI Applications, Ethics & Society
Object detection in images uses:
  1. NLP
  2. Computer Vision
  3. Knowledge representation
  4. Search algorithms
Correct Answer: B) Computer Vision
Question 192
AI Applications, Ethics & Society
Route optimization in Google Maps uses:
  1. Random algorithms
  2. Informed search algorithms
  3. Brute force only
  4. Manual calculation
Correct Answer: B) Informed search algorithms
Question 193
AI Applications, Ethics & Society
Autonomous vehicle navigation uses:
  1. NLP
  2. Knowledge representation only
  3. Computer Vision and AI
  4. Simple rules only
Correct Answer: C) Computer Vision and AI
Question 194
AI Applications, Ethics & Society
Personalized treatment in healthcare uses:
  1. Fixed protocols only
  2. AI and patient data analysis
  3. Random assignment
  4. Manual testing only
Correct Answer: B) AI and patient data analysis
Question 195
AI Applications, Ethics & Society
Customer behavior analysis is AI in:
  1. Healthcare
  2. Education
  3. Business and Marketing
  4. Transportation
Correct Answer: C) Business and Marketing
Question 196
AI Applications, Ethics & Society
Algorithmic bias in AI refers to:
  1. Faster algorithms
  2. Unfair or discriminatory outputs from AI systems
  3. Memory optimization
  4. Coding errors
Correct Answer: B) Unfair or discriminatory outputs from AI systems
Question 197
AI Applications, Ethics & Society
Explainable AI focuses on:
  1. Making AI faster
  2. Making AI decisions transparent and understandable
  3. Reducing AI cost
  4. Expanding AI hardware
Correct Answer: B) Making AI decisions transparent and understandable
Question 198
AI Applications, Ethics & Society
Job displacement is a concern because:
  1. AI creates more jobs always
  2. AI automation may replace human workers
  3. AI is too slow
  4. AI requires human supervision always
Correct Answer: B) AI automation may replace human workers
Question 199
AI Applications, Ethics & Society
Data privacy in AI is a concern because:
  1. AI uses less data
  2. AI systems collect and process large amounts of personal data
  3. AI protects all data
  4. AI ignores user data
Correct Answer: B) AI systems collect and process large amounts of personal data
Question 200
AI Applications, Ethics & Society
Which is a benefit of AI?
  1. Increased costs
  2. Decreased efficiency
  3. Automation of repetitive tasks
  4. Reduced accuracy
Correct Answer: C) Automation of repetitive tasks
Question 201
AI Applications, Ethics & Society
A challenge in AI development is:
  1. Too many solutions
  2. High development cost
  3. Easy implementation
  4. Simple algorithms
Correct Answer: B) High development cost
Question 202
AI Applications, Ethics & Society
Transparency in AI means:
  1. AI systems are visible physically SECTION 10: Advanced, Mixed & Conceptual Questions (Q206–Q500)
  2. AI decision process can be understood
  3. AI code is open source
  4. AI data is public
Correct Answer: B) AI decision process can be understood
Question 203
Advanced, Mixed & Conceptual Questions
Future of AI includes:
  1. AI replacing all humans
  2. Human-AI collaboration
  3. Stopping AI development
  4. AI becoming conscious
Correct Answer: B) Human-AI collaboration
Question 204
Advanced, Mixed & Conceptual Questions
Responsible AI development ensures:
  1. Maximum profit
  2. Fairness, safety and societal benefit
  3. Fastest performance
  4. Minimal features
Correct Answer: B) Fairness, safety and societal benefit
Question 205
Advanced, Mixed & Conceptual Questions
The black box problem in AI refers to:
  1. Computers being black
  2. AI making decisions humans cannot interpret
  3. Encrypted data
  4. Missing hardware
Correct Answer: B) AI making decisions humans cannot interpret
Question 206
Advanced, Mixed & Conceptual Questions
Which search algorithm is complete and optimal with equal costs?
  1. DFS
  2. DLS
  3. BFS
  4. Simple Hill Climbing
Correct Answer: C) BFS
Question 207
Advanced, Mixed & Conceptual Questions
Which search algorithm is NOT complete?
  1. BFS
  2. IDS
  3. DFS
  4. A* with admissible heuristic
Correct Answer: C) DFS
Question 208
Advanced, Mixed & Conceptual Questions
The branching factor 'b' represents:
  1. Depth of tree
  2. Number of child nodes per parent
  3. Memory size
  4. Height of tree
Correct Answer: B) Number of child nodes per parent
Question 209
Advanced, Mixed & Conceptual Questions
IDS is preferred over BFS because:
  1. It is faster
  2. It uses less memory while being complete
  3. It finds longer paths
  4. It uses more memory
Correct Answer: B) It uses less memory while being complete
Question 210
Advanced, Mixed & Conceptual Questions
In A*, if h(n)=0 for all nodes, A* becomes:
  1. DFS
  2. Greedy search
  3. Uniform Cost Search
  4. BFS
Correct Answer: C) Uniform Cost Search
Question 211
Advanced, Mixed & Conceptual Questions
Which heuristic is NOT admissible?
  1. Straight-line distance
  2. Manhattan distance for 8-puzzle
  3. A heuristic that always overestimates
  4. Zero heuristic
Correct Answer: C) A heuristic that always overestimates
Question 212
Advanced, Mixed & Conceptual Questions
In Simulated Annealing, if the move is better (delta E > 0):
  1. Move is rejected
  2. Move is accepted with probability
  3. Move is always accepted
  4. Move is ignored
Correct Answer: C) Move is always accepted
Question 213
Advanced, Mixed & Conceptual Questions
The fitness function in Genetic Algorithm is similar to:
  1. Heuristic function in search
  2. Transition model
  3. Goal test
  4. Initial state
Correct Answer: A) Heuristic function in search
Question 214
Advanced, Mixed & Conceptual Questions
Which statement about local search is TRUE?
  1. It remembers the full path
  2. It only cares about the final state, not the path
  3. It always finds optimal solution
  4. It explores all states
Correct Answer: B) It only cares about the final state, not the path
Question 215
Advanced, Mixed & Conceptual Questions
A* with an inadmissible heuristic:
  1. Always finds optimal path
  2. May not find optimal path
  3. Becomes BFS
  4. Becomes DFS
Correct Answer: B) May not find optimal path
Question 216
Advanced, Mixed & Conceptual Questions
Which agent type explicitly uses a utility function?
  1. Simple reflex agent
  2. Goal-based agent
  3. Utility-based agent
  4. Model-based agent
Correct Answer: C) Utility-based agent
Question 217
Advanced, Mixed & Conceptual Questions
A performance measure for an AI agent measures:
  1. Hardware specifications
  2. How well it achieves its goals
  3. Programming language
  4. Memory size
Correct Answer: B) How well it achieves its goals
Question 218
Advanced, Mixed & Conceptual Questions
The critic in a learning agent:
  1. Makes decisions
  2. Improves the system
  3. Evaluates performance
  4. Suggests actions
Correct Answer: C) Evaluates performance
Question 219
Advanced, Mixed & Conceptual Questions
The problem generator in a learning agent:
  1. Evaluates performance
  2. Makes decisions
  3. Suggests exploratory actions
  4. Stores knowledge
Correct Answer: C) Suggests exploratory actions
Question 220
Advanced, Mixed & Conceptual Questions
Propositional Logic CANNOT represent:
  1. True/false statements
  2. Logical connectives
  3. General statements with variables like all humans
  4. Simple rules
Correct Answer: C) General statements with variables like all humans
Question 221
Advanced, Mixed & Conceptual Questions
ForAll x Bird(x) implies CanFly(x): what breaks this rule?
  1. Eagle
  2. Sparrow
  3. Penguin (cannot fly)
  4. Parrot
Correct Answer: C) Penguin (cannot fly)
Question 222
Advanced, Mixed & Conceptual Questions
Predicate Logic supports:
  1. Only atomic propositions
  2. Quantifiers and predicates over objects
  3. Only numeric values
  4. Only true/false
Correct Answer: B) Quantifiers and predicates over objects
Question 223
Advanced, Mixed & Conceptual Questions
Forward chaining in expert systems is:
  1. Goal-driven
  2. Data-driven
  3. Random-driven
  4. Memory-driven
Correct Answer: B) Data-driven
Question 224
Advanced, Mixed & Conceptual Questions
The PEAS framework is used to:
  1. Design hardware
  2. Define task environment of an agent
  3. Write programs
  4. Sort data
Correct Answer: B) Define task environment of an agent
Question 225
Advanced, Mixed & Conceptual Questions
Backward chaining is used in:
  1. BFS
  2. Forward-chaining expert systems
  3. Prolog programming language
  4. Greedy search
Correct Answer: C) Prolog programming language
Question 226
Advanced, Mixed & Conceptual Questions
Supervised Learning requires:
  1. No data
  2. Labeled training data
  3. Only test data
  4. Random inputs
Correct Answer: B) Labeled training data
Question 227
Advanced, Mixed & Conceptual Questions
Unsupervised Learning works with:
  1. Labeled data
  2. No data
  3. Unlabeled data
  4. Only test data
Correct Answer: C) Unlabeled data
Question 228
Advanced, Mixed & Conceptual Questions
Reinforcement Learning involves:
  1. Only labeled data
  2. Agent learning through rewards and penalties
  3. No feedback
  4. Fixed rules only
Correct Answer: B) Agent learning through rewards and penalties
Question 229
Advanced, Mixed & Conceptual Questions
Neural Networks are inspired by:
  1. Computer circuits
  2. Human brain neurons
  3. Mathematical equations only
  4. Physical laws
Correct Answer: B) Human brain neurons
Question 230
Advanced, Mixed & Conceptual Questions
Deep Learning uses:
  1. Single-layer networks
  2. Multiple layers of neural networks
  3. Only decision trees
  4. Linear algorithms only
Correct Answer: B) Multiple layers of neural networks
Question 231
Advanced, Mixed & Conceptual Questions
Which is NOT a component of uninformed search?
  1. BFS
  2. DFS
  3. A* Search
  4. IDS
Correct Answer: C) A* Search
Question 232
Advanced, Mixed & Conceptual Questions
Which statement about BFS is FALSE?
  1. It uses a queue
  2. It explores level by level
  3. It uses less memory than DFS
  4. It is complete
Correct Answer: C) It uses less memory than DFS
Question 233
Advanced, Mixed & Conceptual Questions
Which statement about DFS is TRUE?
  1. It uses a queue
  2. It guarantees shortest path
  3. It uses less memory than BFS
  4. It explores level by level
Correct Answer: C) It uses less memory than BFS
Question 234
Advanced, Mixed & Conceptual Questions
The main advantage of BFS over DFS is:
  1. Less memory
  2. Faster speed
  3. Guarantees shortest path
  4. More depth
Correct Answer: C) Guarantees shortest path
Question 235
Advanced, Mixed & Conceptual Questions
The main advantage of DFS over BFS is:
  1. Finds shortest path
  2. Less memory usage
  3. More complete
  4. More optimal
Correct Answer: B) Less memory usage
Question 236
Advanced, Mixed & Conceptual Questions
Heuristic search is better than uninformed search because:
  1. Always finds optimal solution
  2. Uses less memory always
  3. Uses domain knowledge to be more efficient
  4. Is simpler
Correct Answer: C) Uses domain knowledge to be more efficient
Question 237
Advanced, Mixed & Conceptual Questions
If h(n) always equals the true cost, A* would:
  1. Expand all nodes
  2. Expand no nodes
  3. Expand only nodes on optimal path
  4. Be same as DFS
Correct Answer: C) Expand only nodes on optimal path
Question 238
Advanced, Mixed & Conceptual Questions
Greedy Best First Search is NOT optimal because:
  1. It uses heuristics
  2. It ignores actual path cost g(n)
  3. It uses priority queue
  4. It explores too many nodes
Correct Answer: B) It ignores actual path cost g(n)
Question 239
Advanced, Mixed & Conceptual Questions
The main advantage of A* over Greedy search is:
  1. A* is faster
  2. A* considers both actual and estimated cost
  3. A* uses less memory
  4. A* never uses heuristics
Correct Answer: B) A* considers both actual and estimated cost
Question 240
Advanced, Mixed & Conceptual Questions
ELIZA is considered an early example of:
  1. Computer Vision
  2. Robotic systems
  3. Conversational AI
  4. Reinforcement Learning
Correct Answer: C) Conversational AI
Question 241
Advanced, Mixed & Conceptual Questions
The main limitation of ELIZA was:
  1. It was too intelligent
  2. It did not truly understand language
  3. It was too slow
  4. It required too much memory
Correct Answer: B) It did not truly understand language
Question 242
Advanced, Mixed & Conceptual Questions
Which is an example of state space?
  1. A database table
  2. All possible configurations of a maze
  3. A sorted array
  4. A linked list
Correct Answer: B) All possible configurations of a maze
Question 243
Advanced, Mixed & Conceptual Questions
Transition model in search shows:
  1. Memory usage
  2. How actions change states
  3. Node depth
  4. Branching factor
Correct Answer: B) How actions change states
Question 244
Advanced, Mixed & Conceptual Questions
Which is NOT a component of state space?
  1. Initial state
  2. Goal test
  3. Compiler
  4. Actions
Correct Answer: C) Compiler
Question 245
Advanced, Mixed & Conceptual Questions
IDS repeats exploration which causes:
  1. Finding wrong paths
  2. Redundant computation at smaller depths
  3. Memory overflow
  4. Incorrect solutions
Correct Answer: B) Redundant computation at smaller depths
Question 246
Advanced, Mixed & Conceptual Questions
John McCarthy is credited with:
  1. Inventing the internet
  2. Coining the term Artificial Intelligence
  3. Proposing the Turing Test
  4. Building ELIZA
Correct Answer: B) Coining the term Artificial Intelligence
Question 247
Advanced, Mixed & Conceptual Questions
Symbolic AI focused on:
  1. Statistical learning
  2. Neural networks
  3. Rule-based and logical reasoning
  4. Deep learning
Correct Answer: C) Rule-based and logical reasoning
Question 248
Advanced, Mixed & Conceptual Questions
An agent's percept sequence is:
  1. Its list of actions
  2. The complete history of observations
  3. Its memory storage
  4. Its goal list
Correct Answer: B) The complete history of observations
Question 249
Advanced, Mixed & Conceptual Questions
A fully observable environment means:
  1. Agent cannot see anything
  2. Agent can perceive the complete state
  3. Agent has partial information
  4. Environment is hidden
Correct Answer: B) Agent can perceive the complete state
Question 250
Advanced, Mixed & Conceptual Questions
A deterministic environment means:
  1. Actions have random effects
  2. Next state fully determined by current state and action
  3. Agent cannot predict outcomes
  4. Multiple possible outcomes
Correct Answer: B) Next state fully determined by current state and action
Question 251
Advanced, Mixed & Conceptual Questions
A stochastic environment has:
  1. Fixed outcomes for actions
  2. Randomness or uncertainty in outcomes
  3. Perfect predictability
  4. No randomness
Correct Answer: B) Randomness or uncertainty in outcomes
Question 252
Advanced, Mixed & Conceptual Questions
A static environment:
  1. Always changes
  2. Changes while agent deliberates
  3. Does not change while agent deliberates
  4. Has no states
Correct Answer: C) Does not change while agent deliberates
Question 253
Advanced, Mixed & Conceptual Questions
A dynamic environment:
  1. Never changes
  2. Changes while the agent is deliberating
  3. Has only two states
  4. Is always observable
Correct Answer: B) Changes while the agent is deliberating
Question 254
Advanced, Mixed & Conceptual Questions
In a competitive multi-agent environment, agents:
  1. Cooperate fully
  2. Compete against each other
  3. Ignore each other
  4. Merge together
Correct Answer: B) Compete against each other
Question 255
Advanced, Mixed & Conceptual Questions
Random restart Hill Climbing solves:
  1. Plateau problem
  2. Local maxima problem by restarting from different points
  3. Ridge problem only
  4. All optimization problems
Correct Answer: B) Local maxima problem by restarting from different points
Question 256
Advanced, Mixed & Conceptual Questions
As Simulated Annealing temperature approaches 0, it becomes like:
  1. BFS
  2. DFS
  3. Hill Climbing
  4. Genetic Algorithm
Correct Answer: C) Hill Climbing
Question 257
Advanced, Mixed & Conceptual Questions
Initial population in Genetic Algorithm is usually:
  1. A single solution
  2. Randomly generated solutions
  3. The optimal solution
  4. Empty
Correct Answer: B) Randomly generated solutions
Question 258
Advanced, Mixed & Conceptual Questions
In Genetic Algorithm, each new generation should be:
  1. Worse than previous
  2. Same as previous
  3. Generally better due to selection and crossover
  4. Random
Correct Answer: C) Generally better due to selection and crossover
Question 259
Advanced, Mixed & Conceptual Questions
Which AI technique mimics Darwinian evolution?
  1. Neural Networks
  2. Hill Climbing
  3. Genetic Algorithms
  4. BFS
Correct Answer: C) Genetic Algorithms
Question 260
Advanced, Mixed & Conceptual Questions
Knowledge acquisition is:
  1. Deleting knowledge
  2. Adding new knowledge to the knowledge base
  3. Sorting knowledge
  4. Compressing knowledge
Correct Answer: B) Adding new knowledge to the knowledge base
Question 261
Advanced, Mixed & Conceptual Questions
Which is an example of inductive reasoning?
  1. Ali is mortal because all humans are mortal
  2. Sun rises tomorrow because it has risen every day
  3. 2+2=4
  4. P implies Q, P therefore Q
Correct Answer: B) Sun rises tomorrow because it has risen every day
Question 262
Advanced, Mixed & Conceptual Questions
Which is an example of deductive reasoning?
  1. Sun rises tomorrow based on observation
  2. All cats are animals, Tom is cat, therefore Tom is animal
  3. Machine learning from data
  4. Estimating distance
Correct Answer: B) All cats are animals, Tom is cat, therefore Tom is animal
Question 263
Advanced, Mixed & Conceptual Questions
Minimax algorithm is used in:
  1. Optimization problems
  2. Two-player adversarial games
  3. Clustering
  4. Image recognition
Correct Answer: B) Two-player adversarial games
Question 264
Advanced, Mixed & Conceptual Questions
Alpha-beta pruning improves minimax by:
  1. Adding more nodes
  2. Eliminating branches that cannot affect final decision
  3. Increasing depth
  4. Adding randomness
Correct Answer: B) Eliminating branches that cannot affect final decision
Question 265
Advanced, Mixed & Conceptual Questions
In game playing, MAX player tries to:
  1. Minimize score
  2. Maximize score
  3. Achieve zero
  4. Make random moves
Correct Answer: B) Maximize score
Question 266
Advanced, Mixed & Conceptual Questions
In game playing, MIN player tries to:
  1. Maximize score
  2. Minimize score
  3. Achieve maximum
  4. Make random moves
Correct Answer: B) Minimize score
Question 267
Advanced, Mixed & Conceptual Questions
Constraint satisfaction problems involve:
  1. Finding any solution
  2. Finding values satisfying all constraints
  3. Random assignment
  4. Sorting constraints
Correct Answer: B) Finding values satisfying all constraints
Question 268
Advanced, Mixed & Conceptual Questions
Sudoku solving is an example of:
  1. Search problem
  2. Constraint satisfaction problem
  3. NLP problem
  4. Computer vision problem
Correct Answer: B) Constraint satisfaction problem
Question 269
Advanced, Mixed & Conceptual Questions
Backtracking in CSP means:
  1. Restarting completely
  2. Going back when constraint is violated
  3. Forward planning
  4. Random restart
Correct Answer: B) Going back when constraint is violated
Question 270
Advanced, Mixed & Conceptual Questions
PROLOG is a programming language based on:
  1. Object-oriented programming
  2. Logic and knowledge representation
  3. Functional programming
  4. Procedural programming
Correct Answer: B) Logic and knowledge representation
Question 271
Advanced, Mixed & Conceptual Questions
Automated theorem proving uses:
  1. Neural networks
  2. Logic and inference mechanisms
  3. Genetic algorithms
  4. Computer vision
Correct Answer: B) Logic and inference mechanisms
Question 272
Advanced, Mixed & Conceptual Questions
AI planning involves:
  1. Only executing actions
  2. Generating sequence of actions to achieve goal
  3. Random action selection
  4. Sorting states
Correct Answer: B) Generating sequence of actions to achieve goal
Question 273
Advanced, Mixed & Conceptual Questions
Which AI field handles speech recognition?
  1. Computer Vision
  2. Robotics
  3. Natural Language Processing
  4. Knowledge representation
Correct Answer: C) Natural Language Processing
Question 274
Advanced, Mixed & Conceptual Questions
Tokenization in NLP is:
  1. Encrypting text
  2. Breaking text into words or tokens
  3. Translating text
  4. Compressing text
Correct Answer: B) Breaking text into words or tokens
Question 275
Advanced, Mixed & Conceptual Questions
Stemming in NLP is:
  1. Adding words to text
  2. Reducing words to their root form
  3. Translating words
  4. Encrypting words
Correct Answer: B) Reducing words to their root form
Question 276
Advanced, Mixed & Conceptual Questions
Sentiment analysis determines:
  1. The topic of text
  2. The language of text
  3. The opinion or emotion in text
  4. The length of text
Correct Answer: C) The opinion or emotion in text
Question 277
Advanced, Mixed & Conceptual Questions
Named Entity Recognition (NER) identifies:
  1. Code variables
  2. Names of people, places, organizations in text
  3. Random words
  4. Sentence structure
Correct Answer: B) Names of people, places, organizations in text
Question 278
Advanced, Mixed & Conceptual Questions
Text classification assigns:
  1. Colors to text
  2. Categories to text documents
  3. Random numbers to words
  4. Fonts to documents
Correct Answer: B) Categories to text documents
Question 279
Advanced, Mixed & Conceptual Questions
Image segmentation divides:
  1. Text into sentences
  2. Image into meaningful regions
  3. Code into functions
  4. Audio into segments
Correct Answer: B) Image into meaningful regions
Question 280
Advanced, Mixed & Conceptual Questions
Convolutional Neural Networks are specialized for:
  1. Text processing
  2. Audio processing
  3. Image processing
  4. Numerical computation only
Correct Answer: C) Image processing
Question 281
Advanced, Mixed & Conceptual Questions
Recurrent Neural Networks are suited for:
  1. Image classification
  2. Sequential data like text and speech
  3. Static data
  4. Sorting
Correct Answer: B) Sequential data like text and speech
Question 282
Advanced, Mixed & Conceptual Questions
Transfer learning means:
  1. Moving data between computers
  2. Reusing a model trained on one task for another
  3. Transferring programs
  4. Moving weights randomly
Correct Answer: B) Reusing a model trained on one task for another
Question 283
Advanced, Mixed & Conceptual Questions
Gradient descent is used to:
  1. Increase error
  2. Minimize loss function in ML
  3. Maximize loss
  4. Sort data
Correct Answer: B) Minimize loss function in ML
Question 284
Advanced, Mixed & Conceptual Questions
Overfitting in ML means:
  1. Model performs well on all data
  2. Model performs well on training but poor on new data
  3. Model is too simple
  4. Model never learns
Correct Answer: B) Model performs well on training but poor on new data
Question 285
Advanced, Mixed & Conceptual Questions
Regularization in ML is used to:
  1. Increase complexity
  2. Prevent overfitting
  3. Speed up training
  4. Increase training data
Correct Answer: B) Prevent overfitting
Question 286
Advanced, Mixed & Conceptual Questions
Decision trees are used in:
  1. Image processing only
  2. Classification and regression
  3. Network routing
  4. Encryption
Correct Answer: B) Classification and regression
Question 287
Advanced, Mixed & Conceptual Questions
Random Forest is:
  1. A single decision tree
  2. An ensemble of multiple decision trees
  3. A clustering algorithm
  4. A search algorithm
Correct Answer: B) An ensemble of multiple decision trees
Question 288
Advanced, Mixed & Conceptual Questions
K-Nearest Neighbors classifies based on:
  1. Decision tree rules
  2. K closest training examples
  3. Mathematical equations
  4. Random selection
Correct Answer: B) K closest training examples
Question 289
Advanced, Mixed & Conceptual Questions
K-Means is an algorithm for:
  1. Classification
  2. Clustering
  3. Regression
  4. Search
Correct Answer: B) Clustering
Question 290
Advanced, Mixed & Conceptual Questions
Naive Bayes is based on:
  1. Decision trees
  2. Bayes theorem with naive independence assumption
  3. Neural networks
  4. Genetic algorithms
Correct Answer: B) Bayes theorem with naive independence assumption
Question 291
Advanced, Mixed & Conceptual Questions
Reinforcement learning agent learns by:
  1. Labeled data
  2. Trial and error with rewards and penalties
  3. Unsupervised clustering
  4. Fixed rules
Correct Answer: B) Trial and error with rewards and penalties
Question 292
Advanced, Mixed & Conceptual Questions
In reinforcement learning, agent aims to:
  1. Minimize cumulative reward
  2. Maximize cumulative reward over time
  3. Achieve zero reward
  4. Ignore environment
Correct Answer: B) Maximize cumulative reward over time
Question 293
Advanced, Mixed & Conceptual Questions
AlphaGo used which AI techniques?
  1. Only search
  2. Only neural networks
  3. Deep learning combined with reinforcement learning and search
  4. Simple rules only
Correct Answer: C) Deep learning combined with reinforcement learning and search
Question 294
Advanced, Mixed & Conceptual Questions
Generative AI can:
  1. Only classify data
  2. Generate new content like text, images, code
  3. Only detect objects
  4. Only sort data
Correct Answer: B) Generate new content like text, images, code
Question 295
Advanced, Mixed & Conceptual Questions
AI hallucination refers to:
  1. AI seeing real things
  2. AI generating plausible but incorrect information
  3. AI becoming conscious
  4. AI processing images
Correct Answer: B) AI generating plausible but incorrect information
Question 296
Advanced, Mixed & Conceptual Questions
Deepfakes are:
  1. Real videos
  2. Fake but realistic-looking AI-generated media
  3. Text documents
  4. Audio only
Correct Answer: B) Fake but realistic-looking AI-generated media
Question 297
Advanced, Mixed & Conceptual Questions
AI in cybersecurity is used for:
  1. Creating malware
  2. Detecting and preventing cyber attacks
  3. Encrypting all data permanently
  4. Removing security
Correct Answer: B) Detecting and preventing cyber attacks
Question 298
Advanced, Mixed & Conceptual Questions
Swarm intelligence is inspired by:
  1. Single experts
  2. Collective behavior of social insects like ants and bees
  3. Mathematical theorems
  4. Physical laws
Correct Answer: B) Collective behavior of social insects like ants and bees
Question 299
Advanced, Mixed & Conceptual Questions
Ant Colony Optimization mimics:
  1. Ant nests building
  2. How ants find shortest paths using pheromones
  3. Ant reproduction
  4. Ant eating behavior
Correct Answer: B) How ants find shortest paths using pheromones
Question 300
Advanced, Mixed & Conceptual Questions
Which is NOT an informed search algorithm?
  1. A*
  2. Greedy Best First Search
  3. Depth First Search
  4. Heuristic search
Correct Answer: C) Depth First Search
Question 301
Advanced, Mixed & Conceptual Questions
The AI winter refers to:
  1. Cold weather affecting computers
  2. Periods of reduced funding and interest in AI
  3. AI systems shutting down
  4. Winter training of AI models
Correct Answer: B) Periods of reduced funding and interest in AI
Question 302
Advanced, Mixed & Conceptual Questions
Explainable AI (XAI) aims to:
  1. Make AI systems larger
  2. Make AI decisions interpretable to humans
  3. Speed up AI
  4. Reduce AI accuracy
Correct Answer: B) Make AI decisions interpretable to humans
Question 303
Advanced, Mixed & Conceptual Questions
Federated learning allows:
  1. Centralized data collection
  2. Training across distributed devices without sharing raw data
  3. Single device training
  4. Public data sharing
Correct Answer: B) Training across distributed devices without sharing raw data
Question 304
Advanced, Mixed & Conceptual Questions
Privacy-preserving AI techniques protect:
  1. Model performance
  2. Individual user data during training
  3. Hardware resources
  4. Network bandwidth
Correct Answer: B) Individual user data during training
Question 305
Advanced, Mixed & Conceptual Questions
AI alignment refers to:
  1. Aligning hardware components
  2. Ensuring AI goals align with human values
  3. Aligning code formatting
  4. Aligning network cables
Correct Answer: B) Ensuring AI goals align with human values
Question 306
Advanced, Mixed & Conceptual Questions
Current AI systems are mostly:
  1. AGI (General AI)
  2. Superintelligent
  3. Narrow AI designed for specific tasks
  4. Conscious systems
Correct Answer: C) Narrow AI designed for specific tasks
Question 307
Advanced, Mixed & Conceptual Questions
AGI stands for:
  1. Advanced Graphics Interface
  2. Artificial General Intelligence
  3. Automated Grid Intelligence
  4. Autonomous Goal Initiative
Correct Answer: B) Artificial General Intelligence
Question 308
Advanced, Mixed & Conceptual Questions
Q-learning is a:
  1. Query language
  2. Reinforcement learning algorithm
  3. Search algorithm
  4. Sorting method
Correct Answer: B) Reinforcement learning algorithm
Question 309
Advanced, Mixed & Conceptual Questions
The backpropagation algorithm is used to:
  1. Forward pass data
  2. Train neural networks by updating weights
  3. Initialize weights only
  4. Classify data
Correct Answer: C) Initialize weights only
Question 310
Advanced, Mixed & Conceptual Questions
Dropout in neural networks is used for:
  1. Removing layers
  2. Regularization to prevent overfitting
  3. Speeding up training
  4. Adding more neurons
Correct Answer: B) Regularization to prevent overfitting
Question 311
Advanced, Mixed & Conceptual Questions
Activation functions in neural networks:
  1. Remove non-linearity
  2. Introduce non-linearity enabling complex learning
  3. Sort outputs
  4. Compress data
Correct Answer: B) Introduce non-linearity enabling complex learning
Question 312
Advanced, Mixed & Conceptual Questions
Sigmoid activation outputs values in range:
  1. -1 to 1
  2. 0 to 1
  3. 0 to infinity
  4. -inf to inf
Correct Answer: B) 0 to 1
Question 313
Advanced, Mixed & Conceptual Questions
Softmax activation is used in:
  1. Binary classification output
  2. Multi-class classification output
  3. Regression output
  4. Input layer
Correct Answer: B) Multi-class classification output
Question 314
Advanced, Mixed & Conceptual Questions
Training data is used to:
  1. Test the model
  2. Fit or train the model parameters
  3. Deploy the model
  4. Evaluate final performance
Correct Answer: B) Fit or train the model parameters
Question 315
Advanced, Mixed & Conceptual Questions
Test data is used to:
  1. Train the model
  2. Validate during training
  3. Evaluate final model performance
  4. Fine-tune parameters
Correct Answer: C) Evaluate final model performance
Question 316
Advanced, Mixed & Conceptual Questions
Hyperparameters in ML are:
  1. Model weights learned from data
  2. Settings configured before training
  3. Training examples
  4. Output predictions
Correct Answer: B) Settings configured before training
Question 317
Advanced, Mixed & Conceptual Questions
Learning rate in ML controls:
  1. How fast computer runs
  2. How much model parameters update each step
  3. Number of layers
  4. Data size
Correct Answer: B) How much model parameters update each step
Question 318
Advanced, Mixed & Conceptual Questions
Epoch in ML training means:
  1. Single data point
  2. One complete pass through training dataset
  3. Single weight update
  4. Model evaluation
Correct Answer: B) One complete pass through training dataset
Question 319
Advanced, Mixed & Conceptual Questions
Which AI concept is used in autonomous planning?
  1. Simple reflex agents
  2. Goal-based and utility-based agents
  3. Databases
  4. Compilers
Correct Answer: B) Goal-based and utility-based agents
Question 320
Advanced, Mixed & Conceptual Questions
Machine Learning is different from traditional programming because:
  1. ML is slower
  2. ML learns patterns from data without explicit rules
  3. ML uses more code
  4. ML requires more hardware
Correct Answer: B) ML learns patterns from data without explicit rules
Question 321
Advanced, Mixed & Conceptual Questions
Which type of ML predicts output from labeled input?
  1. Unsupervised Learning
  2. Reinforcement Learning
  3. Supervised Learning
  4. Deep Learning
Correct Answer: C) Supervised Learning
Question 322
Advanced, Mixed & Conceptual Questions
Which type of ML finds hidden patterns in unlabeled data?
  1. Supervised Learning
  2. Unsupervised Learning
  3. Reinforcement Learning
  4. Transfer Learning
Correct Answer: B) Unsupervised Learning
Question 323
Advanced, Mixed & Conceptual Questions
The concept of rationality in AI means:
  1. Being morally right
  2. Making best decision with available information
  3. Being perfect always
  4. Never making mistakes
Correct Answer: B) Making best decision with available information
Question 324
Advanced, Mixed & Conceptual Questions
LiDAR in self-driving cars is used as:
  1. An actuator
  2. A sensor for distance measurement
  3. A decision system
  4. A display screen
Correct Answer: B) A sensor for distance measurement
Question 325
Advanced, Mixed & Conceptual Questions
MYCIN was a famous expert system for:
  1. Chess playing
  2. Medical diagnosis of blood infections
  3. Route planning
  4. Language translation
Correct Answer: B) Medical diagnosis of blood infections
Question 326
Advanced, Mixed & Conceptual Questions
Particle Swarm Optimization is inspired by:
  1. Particle physics
  2. Flocking behavior of birds and fish
  3. Ant colonies
  4. Genetic evolution
Correct Answer: B) Flocking behavior of birds and fish
Question 327
Advanced, Mixed & Conceptual Questions
Acceptance probability in Simulated Annealing is e^(-dE/T). If temperature T is very high:
  1. Probability of accepting bad move is 0
  2. Probability of accepting bad move is close to 1
  3. Algorithm stops
  4. Only good moves accepted
Correct Answer: B) Probability of accepting bad move is close to 1
Question 328
Advanced, Mixed & Conceptual Questions
Lemmatization in NLP returns:
  1. Random form of word
  2. The dictionary base form of word
  3. Encrypted form
  4. Longest form
Correct Answer: B) The dictionary base form of word
Question 329
Advanced, Mixed & Conceptual Questions
In game theory, Nash Equilibrium refers to:
  1. Maximum utility for one player
  2. State where no player benefits from changing strategy unilaterally
  3. Random outcome
  4. Minimum utility state
Correct Answer: B) State where no player benefits from changing strategy unilaterally
Question 330
Advanced, Mixed & Conceptual Questions
Multi-agent systems involve:
  1. Single learning agent
  2. Multiple autonomous agents interacting
  3. No interaction between agents
  4. Only cooperative agents
Correct Answer: B) Multiple autonomous agents interacting
Question 331
Advanced, Mixed & Conceptual Questions
Emergent behavior in multi-agent systems means:
  1. Predictable behavior
  2. Complex behavior arising from simple agent interactions
  3. Random behavior
  4. Fixed behavior
Correct Answer: B) Complex behavior arising from simple agent interactions
Question 332
Advanced, Mixed & Conceptual Questions
Which is a real-world AI application in agriculture?
  1. Increasing pollution
  2. Crop monitoring and yield prediction
  3. Reducing rainfall
  4. Manual harvesting only
Correct Answer: B) Crop monitoring and yield prediction
Question 333
Advanced, Mixed & Conceptual Questions
AI in manufacturing enables:
  1. Only human labor
  2. Predictive maintenance and quality control
  3. Slower production
  4. Random quality control
Correct Answer: B) Predictive maintenance and quality control
Question 334
Advanced, Mixed & Conceptual Questions
Predictive maintenance uses AI to:
  1. Break equipment
  2. Predict when equipment will fail before it does
  3. Repair broken equipment
  4. Manufacture new equipment
Correct Answer: B) Predict when equipment will fail before it does
Question 335
Advanced, Mixed & Conceptual Questions
GANs (Generative Adversarial Networks) consist of:
  1. Single network
  2. Generator and discriminator networks competing
  3. Three networks
  4. No neural networks
Correct Answer: B) Generator and discriminator networks competing
Question 336
Advanced, Mixed & Conceptual Questions
Transformer architecture revolutionized:
  1. Computer Vision mainly
  2. NLP with attention mechanisms
  3. Robotics only
  4. Search algorithms
Correct Answer: B) NLP with attention mechanisms
Question 337
Advanced, Mixed & Conceptual Questions
BERT is a language model known for:
  1. Generating images
  2. Bidirectional understanding of text context
  3. Playing games
  4. Robotic control
Correct Answer: B) Bidirectional understanding of text context
Question 338
Advanced, Mixed & Conceptual Questions
Fine-tuning a pre-trained model means:
  1. Starting training from scratch
  2. Further training on specific task data
  3. Removing all weights
  4. Reducing model size
Correct Answer: B) Further training on specific task data
Question 339
Advanced, Mixed & Conceptual Questions
AI regulation aims to:
  1. Stop AI development
  2. Ensure AI is developed and deployed responsibly
  3. Increase AI bias
  4. Reduce AI accuracy
Correct Answer: B) Ensure AI is developed and deployed responsibly
Question 340
Advanced, Mixed & Conceptual Questions
Human-in-the-loop AI means:
  1. Humans are replaced by AI
  2. Humans remain involved in AI decision process
  3. AI controls humans
  4. No human involvement
Correct Answer: B) Humans remain involved in AI decision process
Question 341
Advanced, Mixed & Conceptual Questions
Which AI field studies how machines can understand and generate human language?
  1. Computer Vision
  2. Robotics
  3. Natural Language Processing
  4. Search
Correct Answer: C) Natural Language Processing
Question 342
Advanced, Mixed & Conceptual Questions
Object detection involves:
  1. Finding text in images
  2. Identifying and locating objects in images
  3. Colorizing images
  4. Compressing images
Correct Answer: B) Identifying and locating objects in images
Question 343
Advanced, Mixed & Conceptual Questions
Which ML algorithm creates tree-like decision rules?
  1. K-Means
  2. Linear Regression
  3. Decision Tree
  4. Neural Network
Correct Answer: C) Decision Tree
Question 344
Advanced, Mixed & Conceptual Questions
Support Vector Machine (SVM) finds:
  1. Shortest path
  2. Best hyperplane to separate classes
  3. Clusters in data
  4. Decision tree
Correct Answer: B) Best hyperplane to separate classes
Question 345
Advanced, Mixed & Conceptual Questions
Dimensionality reduction reduces:
  1. Model accuracy
  2. Number of input features while preserving information
  3. Training data size
  4. Test accuracy
Correct Answer: B) Number of input features while preserving information
Question 346
Advanced, Mixed & Conceptual Questions
PCA (Principal Component Analysis) is used for:
  1. Classification
  2. Clustering
  3. Dimensionality reduction
  4. Reinforcement learning
Correct Answer: C) Dimensionality reduction
Question 347
Advanced, Mixed & Conceptual Questions
Accuracy in ML classification is:
  1. Number of errors
  2. Percentage of correct predictions
  3. Training time
  4. Model size
Correct Answer: B) Percentage of correct predictions
Question 348
Advanced, Mixed & Conceptual Questions
A confusion matrix shows:
  1. Network architecture
  2. True positives, false positives, true negatives, false negatives
  3. Training history
  4. Model parameters
Correct Answer: B) True positives, false positives, true negatives, false negatives
Question 349
Advanced, Mixed & Conceptual Questions
Collaborative filtering recommends based on:
  1. Item features only
  2. Similar users preferences
  3. Random selection
  4. Alphabetical order
Correct Answer: B) Similar users preferences
Question 350
Advanced, Mixed & Conceptual Questions
Content-based filtering recommends based on:
  1. Other users preferences
  2. Features of items similar to what user liked
  3. Random selection
  4. Most popular items
Correct Answer: B) Features of items similar to what user liked
Question 351
Advanced, Mixed & Conceptual Questions
The Dartmouth Conference was held in:
  1. 1950
  2. 1956
  3. 1960
  4. 1970
Correct Answer: B) 1956
Question 352
Advanced, Mixed & Conceptual Questions
Alan Turing published his paper on machine intelligence in:
  1. 1945
  2. 1950
  3. 1956
  4. 1960
Correct Answer: B) 1950
Question 353
Advanced, Mixed & Conceptual Questions
Expert systems expert knowledge comes from:
  1. Databases only
  2. Human experts encoded in rules
  3. Neural networks
  4. Random generation
Correct Answer: B) Human experts encoded in rules
Question 354
Advanced, Mixed & Conceptual Questions
Alpha-beta pruning is a type of:
  1. Game creation
  2. Search optimization for game trees
  3. Neural network training
  4. Data sorting
Correct Answer: B) Search optimization for game trees
Question 355
Advanced, Mixed & Conceptual Questions
AI planning requires:
  1. Only current state knowledge
  2. Initial state, goal state, and available actions
  3. Only goal state
  4. Only actions list
Correct Answer: B) Initial state, goal state, and available actions
Question 356
Advanced, Mixed & Conceptual Questions
Arc consistency in CSP ensures:
  1. All arcs are removed
  2. Values in domains are consistent with neighboring constraints
  3. Random arc assignment
  4. All values removed
Correct Answer: B) Values in domains are consistent with neighboring constraints
Question 357
Advanced, Mixed & Conceptual Questions
AlphaFold is an AI system that:
  1. Plays chess
  2. Predicts protein 3D structures
  3. Drives cars
  4. Translates languages
Correct Answer: B) Predicts protein 3D structures
Question 358
Advanced, Mixed & Conceptual Questions
Text-to-speech (TTS) systems convert:
  1. Speech to text
  2. Text to audio speech
  3. Text to images
  4. Audio to images
Correct Answer: B) Text to audio speech
Question 359
Advanced, Mixed & Conceptual Questions
Prompt engineering is:
  1. Building physical engines
  2. Crafting effective inputs to guide AI model responses
  3. Writing code only
  4. Hardware configuration
Correct Answer: B) Crafting effective inputs to guide AI model responses
Question 360
Advanced, Mixed & Conceptual Questions
Zero-shot learning means:
  1. Training on zero data
  2. Model can handle tasks without task-specific training
  3. Model fails all tasks
  4. Model resets weights
Correct Answer: B) Model can handle tasks without task-specific training
Question 361
Advanced, Mixed & Conceptual Questions
Which type of ML is used in game-playing AI like chess?
  1. Supervised Learning
  2. Unsupervised Learning
  3. Reinforcement Learning
  4. Transfer Learning
Correct Answer: C) Reinforcement Learning
Question 362
Advanced, Mixed & Conceptual Questions
Batch size in ML refers to:
  1. Total dataset size
  2. Number of samples processed in one training step
  3. Model size
  4. Number of epochs
Correct Answer: B) Number of samples processed in one training step
Question 363
Advanced, Mixed & Conceptual Questions
L1 regularization adds penalty based on:
  1. Square of weights
  2. Absolute value of weights
  3. Number of layers
  4. Learning rate
Correct Answer: B) Absolute value of weights
Question 364
Advanced, Mixed & Conceptual Questions
L2 regularization adds penalty based on:
  1. Absolute value of weights
  2. Square of weights
  3. Number of layers
  4. Batch size
Correct Answer: B) Square of weights
Question 365
Advanced, Mixed & Conceptual Questions
ReLU activation function outputs:
  1. Always 0 or 1
  2. max(0,x) – zero for negative, x for positive
  3. Always negative
  4. Random values
Correct Answer: B) max(0,x) – zero for negative, x for positive
Question 366
Advanced, Mixed & Conceptual Questions
Which field studies how agents interact in economic scenarios using AI?
  1. Computer Vision
  2. Game Theory in AI
  3. NLP
  4. Robotics only
Correct Answer: B) Game Theory in AI
Question 367
Advanced, Mixed & Conceptual Questions
Semantic networks represent:
  1. Network hardware
  2. Relationships between concepts as nodes and links
  3. File systems
  4. Database schemas
Correct Answer: B) Relationships between concepts as nodes and links
Question 368
Advanced, Mixed & Conceptual Questions
A knowledge base rule stores:
  1. Hardware configuration
  2. Facts and if-then rules for inference
  3. Network topology
  4. Program code
Correct Answer: B) Facts and if-then rules for inference
Question 369
Advanced, Mixed & Conceptual Questions
Which AI technique is best for large combinatorial optimization?
  1. BFS
  2. Simple Hill Climbing
  3. Genetic Algorithms
  4. DLS
Correct Answer: C) Genetic Algorithms
Question 370
Advanced, Mixed & Conceptual Questions
An ontology in AI defines:
  1. Hardware components
  2. Concepts and relationships in a domain
  3. Network protocols
  4. Programming syntax
Correct Answer: B) Concepts and relationships in a domain
Question 371
Advanced, Mixed & Conceptual Questions
Map coloring is a classic example of:
  1. Image processing
  2. Constraint satisfaction
  3. Search without constraints
  4. NLP
Correct Answer: B) Constraint satisfaction
Question 372
Advanced, Mixed & Conceptual Questions
Which of the following best describes the difference between informed and uninformed search?
  1. Informed search uses less memory
  2. Informed search uses domain knowledge (heuristics) to guide search
  3. Uninformed search is always optimal
  4. They are the same
Correct Answer: B) Informed search uses domain knowledge (heuristics) to guide search
Question 373
Advanced, Mixed & Conceptual Questions
In the 8-puzzle problem, what does each state represent?
  1. A specific tile color
  2. A specific arrangement of tiles on the board
  3. A path through the grid
  4. A goal position
Correct Answer: B) A specific arrangement of tiles on the board
Question 374
Advanced, Mixed & Conceptual Questions
Which data structure does BFS use and why?
  1. Stack, because it explores deepest first
  2. Queue (FIFO), because it explores shallowest nodes first
  3. Heap, for priority ordering
  4. Array, for random access
Correct Answer: B) Queue (FIFO), because it explores shallowest nodes first
Question 375
Advanced, Mixed & Conceptual Questions
Which data structure does DFS use and why?
  1. Queue, because it explores level by level
  2. Stack (LIFO), because it goes deep before backtracking
  3. Heap, for ordering
  4. Array, for indexing
Correct Answer: B) Stack (LIFO), because it goes deep before backtracking
Question 376
Advanced, Mixed & Conceptual Questions
IDS avoids the memory problem of BFS by:
  1. Using a smaller queue
  2. Not storing all frontier nodes, using DFS-style stack
  3. Skipping nodes
  4. Using compression
Correct Answer: B) Not storing all frontier nodes, using DFS-style stack
Question 377
Advanced, Mixed & Conceptual Questions
The evaluation function f(n)=g(n)+h(n) balances:
  1. Speed and memory
  2. Actual path cost and estimated remaining cost
  3. Depth and breadth
  4. Number of nodes and edges
Correct Answer: B) Actual path cost and estimated remaining cost
Question 378
Advanced, Mixed & Conceptual Questions
Why is A* better than Greedy search in most cases?
  1. A* is always faster
  2. A* considers actual path cost preventing poor paths
  3. A* uses less memory
  4. A* uses simpler heuristics
Correct Answer: B) A* considers actual path cost preventing poor paths
Question 379
Advanced, Mixed & Conceptual Questions
What is the role of temperature in Simulated Annealing?
  1. Controls CPU speed
  2. Controls probability of accepting worse solutions
  3. Controls number of iterations
  4. Controls memory usage
Correct Answer: B) Controls probability of accepting worse solutions
Question 380
Advanced, Mixed & Conceptual Questions
How does Genetic Algorithm avoid local optima?
  1. By always choosing best neighbor
  2. By working with population and using crossover and mutation
  3. By using temperature parameter
  4. By restarting randomly
Correct Answer: B) By working with population and using crossover and mutation
Question 381
Advanced, Mixed & Conceptual Questions
Crossover in Genetic Algorithm is analogous to:
  1. Mutation in biology
  2. Biological reproduction combining parental traits
  3. Natural death
  4. Random mutation only
Correct Answer: B) Biological reproduction combining parental traits
Question 382
Advanced, Mixed & Conceptual Questions
Which AI reasoning is used in rule-based expert systems?
  1. Inductive reasoning
  2. Probabilistic reasoning
  3. Deductive reasoning
  4. Statistical reasoning
Correct Answer: C) Deductive reasoning
Question 383
Advanced, Mixed & Conceptual Questions
Which reasoning is the basis of Machine Learning?
  1. Deductive reasoning
  2. Mathematical proof
  3. Inductive reasoning from data
  4. Backward chaining
Correct Answer: C) Inductive reasoning from data
Question 384
Advanced, Mixed & Conceptual Questions
Knowledge representation is needed because:
  1. AI systems need data storage
  2. AI systems need structured information to reason intelligently
  3. AI systems need hardware
  4. AI systems need networks
Correct Answer: B) AI systems need structured information to reason intelligently
Question 385
Advanced, Mixed & Conceptual Questions
Why is Predicate Logic preferred over Propositional Logic in AI?
  1. It is simpler
  2. It can represent objects, relationships and quantifiers
  3. It uses less memory
  4. It is faster
Correct Answer: B) It can represent objects, relationships and quantifiers
Question 386
Advanced, Mixed & Conceptual Questions
The existential quantifier is needed to express:
  1. Universal truths
  2. Statements about all elements
  3. Statements about at least one element
  4. False statements
Correct Answer: C) Statements about at least one element
Question 387
Advanced, Mixed & Conceptual Questions
The universal quantifier is needed to express:
  1. Statements about specific elements
  2. Statements that apply to all elements in domain
  3. Probability statements
  4. Existential claims
Correct Answer: B) Statements that apply to all elements in domain
Question 388
Advanced, Mixed & Conceptual Questions
Forward chaining is best when:
  1. Goal is known and you verify conditions
  2. Facts are known and you want to derive conclusions
  3. You have no facts
  4. Goals are unknown
Correct Answer: B) Facts are known and you want to derive conclusions
Question 389
Advanced, Mixed & Conceptual Questions
Backward chaining is best when:
  1. Facts are known and you derive conclusions
  2. Specific goal is known and you verify if facts support it
  3. You have no goals
  4. Facts are random
Correct Answer: B) Specific goal is known and you verify if facts support it
Question 390
Advanced, Mixed & Conceptual Questions
ELIZA's main technique was pattern matching which means:
  1. It understood language deeply
  2. It matched input patterns to templates without understanding
  3. It used neural networks
  4. It used deep learning
Correct Answer: B) It matched input patterns to templates without understanding
Question 391
Advanced, Mixed & Conceptual Questions
Why is admissibility important for A*?
  1. It makes search faster
  2. It guarantees that A* finds the optimal solution
  3. It reduces memory usage
  4. It simplifies heuristic computation
Correct Answer: B) It guarantees that A* finds the optimal solution
Question 392
Advanced, Mixed & Conceptual Questions
What happens if depth limit in DLS is too small?
  1. Algorithm finds wrong solution
  2. Solution may not be found if it is deeper than limit
  3. Algorithm runs faster
  4. Memory overflows
Correct Answer: B) Solution may not be found if it is deeper than limit
Question 393
Advanced, Mixed & Conceptual Questions
Why does Hill Climbing fail on ridges?
  1. It cannot compute neighbors
  2. The optimal path requires moves that appear worse initially
  3. It has no memory
  4. It uses wrong heuristic
Correct Answer: B) The optimal path requires moves that appear worse initially
Question 394
Advanced, Mixed & Conceptual Questions
How does Genetic Algorithm represent solutions?
  1. As graphs
  2. As chromosomes (strings of genes)
  3. As trees
  4. As matrices
Correct Answer: B) As chromosomes (strings of genes)
Question 395
Advanced, Mixed & Conceptual Questions
What makes a multi-agent environment complex?
  1. More nodes
  2. Agent actions affect other agents creating interdependencies
  3. Larger state space only
  4. More memory needed
Correct Answer: B) Agent actions affect other agents creating interdependencies
Question 396
Advanced, Mixed & Conceptual Questions
AI systems are said to learn when they:
  1. Follow fixed rules
  2. Improve performance on tasks through experience without explicit programming
  3. Execute predetermined steps
  4. Follow database queries
Correct Answer: B) Improve performance on tasks through experience without explicit programming
Question 397
Advanced, Mixed & Conceptual Questions
The Turing Test is NOT a test of:
  1. Human-like conversation
  2. Machine intelligence simulation
  3. True machine consciousness or understanding
  4. Response timing
Correct Answer: C) True machine consciousness or understanding
Question 398
Advanced, Mixed & Conceptual Questions
IBM Watson won Jeopardy using:
  1. Simple keyword matching
  2. AI, NLP, and knowledge retrieval systems
  3. Random guessing
  4. Fixed rules only
Correct Answer: B) AI, NLP, and knowledge retrieval systems
Question 399
Advanced, Mixed & Conceptual Questions
In AI, an action is defined as:
  1. A state of the system
  2. A choice available to the agent that causes state transitions
  3. A data structure
  4. A memory location
Correct Answer: B) A choice available to the agent that causes state transitions
Question 400
Advanced, Mixed & Conceptual Questions
The efficiency of A* depends mainly on:
  1. Hardware speed
  2. Quality of heuristic function
  3. Depth of goal
  4. Memory size
Correct Answer: B) Quality of heuristic function
Question 401
Advanced, Mixed & Conceptual Questions
A state in search problem represents:
  1. An algorithm step
  2. A configuration of the world at a given moment
  3. A data structure
  4. A memory address
Correct Answer: B) A configuration of the world at a given moment
Question 402
Advanced, Mixed & Conceptual Questions
Goal state in search is:
  1. The starting point
  2. Any intermediate state
  3. The desired final configuration
  4. A random state
Correct Answer: C) The desired final configuration
Question 403
Advanced, Mixed & Conceptual Questions
The solution to a search problem is:
  1. A heuristic value
  2. A sequence of actions leading from initial to goal state
  3. A single action
  4. A database entry
Correct Answer: B) A sequence of actions leading from initial to goal state
Question 404
Advanced, Mixed & Conceptual Questions
Optimal solution in search means:
  1. First solution found
  2. Solution with lowest path cost
  3. Longest solution
  4. Random solution
Correct Answer: B) Solution with lowest path cost
Question 405
Advanced, Mixed & Conceptual Questions
Which is NOT an advantage of Genetic Algorithms?
  1. Explores large search space
  2. Avoids local optima
  3. Always faster than Hill Climbing
  4. Good for complex optimization
Correct Answer: C) Always faster than Hill Climbing
Question 406
Advanced, Mixed & Conceptual Questions
The population size in Genetic Algorithm affects:
  1. Only speed
  2. Both diversity and computational cost
  3. Only memory
  4. Only accuracy
Correct Answer: B) Both diversity and computational cost
Question 407
Advanced, Mixed & Conceptual Questions
In Hill Climbing, a shoulder (flat region) is similar to:
  1. A local maximum
  2. A plateau
  3. A global maximum
  4. A ridge
Correct Answer: B) A plateau
Question 408
Advanced, Mixed & Conceptual Questions
Which search explores nodes in order of f(n)=g(n)+h(n)?
  1. BFS
  2. DFS
  3. A*
  4. DLS
Correct Answer: C) A*
Question 409
Advanced, Mixed & Conceptual Questions
Uniform Cost Search is like A* with:
  1. h(n) = infinity
  2. h(n) = 0
  3. g(n) = 0
  4. f(n) = 0
Correct Answer: B) h(n) = 0
Question 410
Advanced, Mixed & Conceptual Questions
What is the main goal of an AI system?
  1. To replace humans completely
  2. To perform tasks that require human-level intelligence
  3. To only process data
  4. To manage hardware
Correct Answer: B) To perform tasks that require human-level intelligence
Question 411
Advanced, Mixed & Conceptual Questions
Knowledge engineering is the process of:
  1. Building hardware
  2. Eliciting and encoding expert knowledge into AI systems
  3. Designing networks
  4. Writing test cases
Correct Answer: B) Eliciting and encoding expert knowledge into AI systems
Question 412
Advanced, Mixed & Conceptual Questions
The frame problem in AI refers to:
  1. Display resolution issues
  2. Difficulty of representing what does NOT change when actions occur
  3. Memory frames
  4. Video frames
Correct Answer: B) Difficulty of representing what does NOT change when actions occur
Question 413
Advanced, Mixed & Conceptual Questions
Fuzzy logic handles:
  1. Only true/false values
  2. Degrees of truth between 0 and 1
  3. Only binary values
  4. Only integers
Correct Answer: B) Degrees of truth between 0 and 1
Question 414
Advanced, Mixed & Conceptual Questions
Bayesian networks represent:
  1. Social networks
  2. Probabilistic relationships between variables
  3. Neural connections
  4. Computer networks
Correct Answer: B) Probabilistic relationships between variables
Question 415
Advanced, Mixed & Conceptual Questions
Natural language understanding (NLU) in AI is harder than generation because:
  1. It requires less processing
  2. Language has ambiguity, context-dependence and implicit meaning
  3. It is well-solved
  4. It only needs pattern matching
Correct Answer: B) Language has ambiguity, context-dependence and implicit meaning
Question 416
Advanced, Mixed & Conceptual Questions
Which of the following is a structured knowledge representation?
  1. Random text
  2. Semantic networks and ontologies
  3. Binary data
  4. Raw sensor data
Correct Answer: B) Semantic networks and ontologies
Question 417
Advanced, Mixed & Conceptual Questions
The closed world assumption in AI means:
  1. World is always changing
  2. What is not known to be true is assumed false
  3. All facts are unknown
  4. Only open questions matter
Correct Answer: B) What is not known to be true is assumed false
Question 418
Advanced, Mixed & Conceptual Questions
An AI agent is autonomous because:
  1. It needs constant human input
  2. It makes its own decisions based on perceptions and goals
  3. It only follows commands
  4. It requires external power only
Correct Answer: B) It makes its own decisions based on perceptions and goals
Question 419
Advanced, Mixed & Conceptual Questions
Which is an example of a partially observable environment?
  1. Chess (perfect info)
  2. Poker (hidden cards)
  3. Tic-tac-toe
  4. Checkers
Correct Answer: B) Poker (hidden cards)
Question 420
Advanced, Mixed & Conceptual Questions
Which is an example of a fully observable environment?
  1. Poker with hidden cards
  2. Chess where all pieces are visible
  3. Autonomous driving
  4. Real-world robotics
Correct Answer: B) Chess where all pieces are visible
Question 421
Advanced, Mixed & Conceptual Questions
The AI field of computer vision primarily processes:
  1. Audio signals
  2. Text data
  3. Visual data like images and video
  4. Numerical data only
Correct Answer: C) Visual data like images and video
Question 422
Advanced, Mixed & Conceptual Questions
Facial recognition uses which AI technique?
  1. NLP
  2. Deep learning-based Computer Vision
  3. Search algorithms
  4. Decision trees only
Correct Answer: B) Deep learning-based Computer Vision
Question 423
Advanced, Mixed & Conceptual Questions
Medical imaging AI is used for:
  1. Administrative tasks
  2. Detecting diseases in scans like X-rays and MRIs
  3. Patient scheduling
  4. Hospital management
Correct Answer: B) Detecting diseases in scans like X-rays and MRIs
Question 424
Advanced, Mixed & Conceptual Questions
Autonomous robots use AI to:
  1. Only follow fixed paths
  2. Navigate, perceive and act in environments intelligently
  3. Only perform repetitive tasks
  4. Only operate under human control
Correct Answer: B) Navigate, perceive and act in environments intelligently
Question 425
Advanced, Mixed & Conceptual Questions
Which is a challenge in developing autonomous vehicles?
  1. Engine performance
  2. Handling all edge cases and unpredictable human behavior
  3. Fuel efficiency
  4. Road construction
Correct Answer: B) Handling all edge cases and unpredictable human behavior
Question 426
Advanced, Mixed & Conceptual Questions
AI in education provides:
  1. Fixed curriculum only
  2. Personalized and adaptive learning experiences
  3. No assessment
  4. Only video lectures
Correct Answer: B) Personalized and adaptive learning experiences
Question 427
Advanced, Mixed & Conceptual Questions
The general problem solver (GPS) was an early AI program designed to:
  1. Calculate GPS coordinates
  2. Solve any problem using means-ends analysis
  3. Play chess only
  4. Process images
Correct Answer: B) Solve any problem using means-ends analysis
Question 428
Advanced, Mixed & Conceptual Questions
Means-ends analysis in AI involves:
  1. Starting from random state
  2. Reducing difference between current state and goal state
  3. Ignoring the goal
  4. Random problem solving
Correct Answer: B) Reducing difference between current state and goal state
Question 429
Advanced, Mixed & Conceptual Questions
MACSYMA was an early AI system for:
  1. Medical diagnosis
  2. Symbolic mathematics
  3. Game playing
  4. Language translation
Correct Answer: B) Symbolic mathematics
Question 430
Advanced, Mixed & Conceptual Questions
Which problem did STUDENT (1964 AI program) solve?
  1. Chess problems
  2. Algebra word problems
  3. Image recognition
  4. Speech recognition
Correct Answer: B) Algebra word problems
Question 431
Advanced, Mixed & Conceptual Questions
Case-based reasoning in AI solves new problems by:
  1. Applying fixed rules
  2. Using solutions from similar past cases
  3. Random guessing
  4. Neural network training
Correct Answer: B) Using solutions from similar past cases
Question 432
Advanced, Mixed & Conceptual Questions
Analogical reasoning finds solutions by:
  1. Applying general rules
  2. Finding similar problems and adapting their solutions
  3. Random search
  4. Exhaustive enumeration
Correct Answer: B) Finding similar problems and adapting their solutions
Question 433
Advanced, Mixed & Conceptual Questions
Which AI technique uses evolutionary principles for optimization?
  1. BFS
  2. DFS
  3. Genetic Algorithms
  4. Backward chaining
Correct Answer: C) Genetic Algorithms
Question 434
Advanced, Mixed & Conceptual Questions
The no free lunch theorem in AI/ML states:
  1. Best algorithm exists for all problems
  2. No algorithm is best for all problems
  3. All algorithms perform equally on all problems
  4. Simple algorithms always win
Correct Answer: B) No algorithm is best for all problems
Question 435
Advanced, Mixed & Conceptual Questions
Which is the most memory-intensive uninformed search?
  1. DFS
  2. DLS
  3. BFS
  4. IDS
Correct Answer: C) BFS
Question 436
Advanced, Mixed & Conceptual Questions
Time complexity of IDS is same as BFS meaning:
  1. IDS is exactly as fast as BFS
  2. Both are O(b^d) though IDS repeats some work
  3. IDS is always faster
  4. IDS is always slower
Correct Answer: B) Both are O(b^d) though IDS repeats some work
Question 437
Advanced, Mixed & Conceptual Questions
Which would you choose for a very deep search tree?
  1. BFS (too much memory)
  2. IDS (memory efficient and complete)
  3. DLS with wrong limit
  4. Random search
Correct Answer: B) IDS (memory efficient and complete)
Question 438
Advanced, Mixed & Conceptual Questions
Heuristic function quality determines:
  1. Algorithm correctness
  2. How many nodes A* needs to expand
  3. Data structure used
  4. Goal depth
Correct Answer: B) How many nodes A* needs to expand
Question 439
Advanced, Mixed & Conceptual Questions
An inadmissible heuristic might:
  1. Always find optimal solution
  2. Cause A* to find suboptimal solution
  3. Make A* faster and optimal
  4. Have no effect
Correct Answer: B) Cause A* to find suboptimal solution
Question 440
Advanced, Mixed & Conceptual Questions
The state space landscape in local search shows:
  1. Physical terrain of problems
  2. Objective function values over all states
  3. Memory usage
  4. Network topology
Correct Answer: B) Objective function values over all states
Question 441
Advanced, Mixed & Conceptual Questions
Which factor most affects Simulated Annealing performance?
  1. Branching factor
  2. Cooling schedule (how temperature decreases)
  3. Initial state only
  4. Memory size
Correct Answer: B) Cooling schedule (how temperature decreases)
Question 442
Advanced, Mixed & Conceptual Questions
Genetic Algorithm's crossover point is selected:
  1. Always at middle
  2. Randomly or based on strategy
  3. At the end
  4. At the beginning always
Correct Answer: B) Randomly or based on strategy
Question 443
Advanced, Mixed & Conceptual Questions
Which AI technique can solve the traveling salesman problem approximately?
  1. BFS
  2. Exact enumeration
  3. Genetic Algorithms or Simulated Annealing
  4. DFS
Correct Answer: C) Genetic Algorithms or Simulated Annealing
Question 444
Advanced, Mixed & Conceptual Questions
The traveling salesman problem is a:
  1. Simple sorting problem
  2. Complex combinatorial optimization problem
  3. Graph coloring problem
  4. Constraint satisfaction problem
Correct Answer: B) Complex combinatorial optimization problem
Question 445
Advanced, Mixed & Conceptual Questions
In AI search, optimality means:
  1. Finding any solution
  2. Finding the best solution with lowest cost
  3. Finding solution fastest
  4. Finding solution with least memory
Correct Answer: B) Finding the best solution with lowest cost
Question 446
Advanced, Mixed & Conceptual Questions
Completeness in AI search means:
  1. Exploring all possible solutions
  2. Guaranteeing to find solution if one exists
  3. Finding optimal solution
  4. Using minimal memory
Correct Answer: B) Guaranteeing to find solution if one exists
Question 447
Advanced, Mixed & Conceptual Questions
Time complexity in AI search measures:
  1. Clock time in seconds
  2. Number of nodes generated during search
  3. Memory in bytes
  4. Code lines
Correct Answer: B) Number of nodes generated during search
Question 448
Advanced, Mixed & Conceptual Questions
Space complexity in AI search measures:
  1. Physical storage
  2. Maximum number of nodes stored in memory
  3. Number of states
  4. Time elapsed
Correct Answer: B) Maximum number of nodes stored in memory
Question 449
Advanced, Mixed & Conceptual Questions
Which is a TRUE statement about uninformed vs informed search?
  1. Uninformed is always better
  2. Informed search uses heuristics making it generally more efficient
  3. They are identical in practice
  4. Uninformed always uses less memory
Correct Answer: B) Informed search uses heuristics making it generally more efficient
Question 450
Advanced, Mixed & Conceptual Questions
A star search guarantees optimal solution when heuristic is:
  1. Overestimating
  2. Admissible (never overestimating)
  3. Always zero
  4. Maximum possible
Correct Answer: B) Admissible (never overestimating)
Question 451
Advanced, Mixed & Conceptual Questions
Which AI application uses search algorithms directly?
  1. Image classification
  2. Text generation
  3. Puzzle solving like 8-puzzle
  4. Speech recognition
Correct Answer: C) Puzzle solving like 8-puzzle
Question 452
Advanced, Mixed & Conceptual Questions
In knowledge representation, a frame represents:
  1. Display resolution
  2. Stereotyped situation with slots for attributes
  3. Video frame
  4. Memory frame
Correct Answer: B) Stereotyped situation with slots for attributes
Question 453
Advanced, Mixed & Conceptual Questions
Scripts in AI knowledge representation capture:
  1. Programming scripts
  2. Stereotyped sequences of events like restaurant visit
  3. Shell scripts
  4. Test scripts
Correct Answer: B) Stereotyped sequences of events like restaurant visit
Question 454
Advanced, Mixed & Conceptual Questions
Propositional calculus is equivalent to:
  1. Predicate Logic
  2. Propositional Logic
  3. Fuzzy Logic
  4. Temporal Logic
Correct Answer: B) Propositional Logic
Question 455
Advanced, Mixed & Conceptual Questions
First-Order Logic is equivalent to:
  1. Propositional Logic
  2. Predicate Logic
  3. Fuzzy Logic
  4. Boolean Logic
Correct Answer: B) Predicate Logic
Question 456
Advanced, Mixed & Conceptual Questions
Which reasoning combines general rule and specific fact for conclusion?
  1. Inductive reasoning
  2. Deductive reasoning (Modus Ponens)
  3. Abductive reasoning
  4. Statistical reasoning
Correct Answer: B) Deductive reasoning (Modus Ponens)
Question 457
Advanced, Mixed & Conceptual Questions
Abductive reasoning finds:
  1. Guaranteed conclusions
  2. Most plausible explanation for observations
  3. General rules from data
  4. Random conclusions
Correct Answer: B) Most plausible explanation for observations
Question 458
Advanced, Mixed & Conceptual Questions
Which is NOT a type of reasoning used in AI?
  1. Deductive
  2. Inductive
  3. Abductive
  4. Circular
Correct Answer: D) Circular
Question 459
Advanced, Mixed & Conceptual Questions
Probabilistic reasoning uses:
  1. Only true or false values
  2. Probability values to handle uncertainty
  3. Only binary logic
  4. Fixed rules only
Correct Answer: B) Probability values to handle uncertainty
Question 460
Advanced, Mixed & Conceptual Questions
Bayesian reasoning updates beliefs based on:
  1. Fixed rules
  2. New evidence using Bayes theorem
  3. Random changes
  4. Hardware signals
Correct Answer: B) New evidence using Bayes theorem
Question 461
Advanced, Mixed & Conceptual Questions
Which search algorithm explores all nodes at depth d before depth d+1?
  1. DFS
  2. BFS
  3. DLS
  4. Hill Climbing
Correct Answer: B) BFS
Question 462
Advanced, Mixed & Conceptual Questions
Which search algorithm explores all nodes at depth d before d+1 using O(bd) memory?
  1. BFS
  2. DFS
  3. IDS
  4. A*
Correct Answer: C) IDS
Question 463
Advanced, Mixed & Conceptual Questions
If branching factor b=2 and goal depth d=3, BFS explores at most how many nodes?
  1. 6
  2. 8
  3. b^(d+1) – 1 = 15
  4. 4
Correct Answer: C) b^(d+1) – 1 = 15
Question 464
Advanced, Mixed & Conceptual Questions
DFS worst case explores how many nodes for tree with depth m?
  1. b^m nodes
  2. b*m nodes
  3. m nodes
  4. b nodes
Correct Answer: A) b^m nodes
Question 465
Advanced, Mixed & Conceptual Questions
In A*, which node is always expanded next?
  1. Deepest node
  2. Node with lowest f(n) = g(n) + h(n)
  3. Node with lowest g(n)
  4. Random node
Correct Answer: B) Node with lowest f(n) = g(n) + h(n)
Question 466
Advanced, Mixed & Conceptual Questions
Greedy Best First Search uses priority queue ordered by:
  1. g(n)
  2. f(n) = g(n)+h(n)
  3. h(n)
  4. Random order
Correct Answer: C) h(n)
Question 467
Advanced, Mixed & Conceptual Questions
Which search is used when path does not matter, only final state?
  1. BFS
  2. DFS
  3. Local search
  4. A*
Correct Answer: C) Local search
Question 468
Advanced, Mixed & Conceptual Questions
Hill Climbing is like DFS but:
  1. Uses queue
  2. Only keeps track of best neighbor, not full tree
  3. Explores level by level
  4. Uses heuristics like A*
Correct Answer: B) Only keeps track of best neighbor, not full tree
Question 469
Advanced, Mixed & Conceptual Questions
Genetic Algorithm stops when:
  1. Memory runs out
  2. Best fitness reaches threshold or max generations reached
  3. Time expires only
  4. Population becomes 0
Correct Answer: B) Best fitness reaches threshold or max generations reached
Question 470
Advanced, Mixed & Conceptual Questions
The 'survival of the fittest' in Genetic Algorithm corresponds to:
  1. Mutation step
  2. Crossover step
  3. Selection step based on fitness
  4. Initialization step
Correct Answer: C) Selection step based on fitness
Question 471
Advanced, Mixed & Conceptual Questions
Which AI concept is demonstrated by ELIZA despite its simplicity?
  1. Deep language understanding
  2. Humans attribute intelligence to systems that mimic conversation
  3. Consciousness in machines
  4. Perfect reasoning
Correct Answer: B) Humans attribute intelligence to systems that mimic conversation
Question 472
Advanced, Mixed & Conceptual Questions
Intelligent agents in multi-agent systems can be:
  1. Only cooperative
  2. Cooperative or competitive depending on goals
  3. Only competitive
  4. Always isolated
Correct Answer: B) Cooperative or competitive depending on goals
Question 473
Advanced, Mixed & Conceptual Questions
Which environment type has actions with unpredictable outcomes?
  1. Deterministic
  2. Stochastic
  3. Observable
  4. Static
Correct Answer: B) Stochastic
Question 474
Advanced, Mixed & Conceptual Questions
For a chess-playing AI, the environment is:
  1. Stochastic and partially observable
  2. Deterministic and fully observable
  3. Continuous and dynamic
  4. Random and unobservable
Correct Answer: B) Deterministic and fully observable
Question 475
Advanced, Mixed & Conceptual Questions
For an autonomous car, the environment is:
  1. Deterministic and fully observable
  2. Stochastic, partially observable, continuous and dynamic
  3. Static and discrete
  4. Fully predictable
Correct Answer: B) Stochastic, partially observable, continuous and dynamic
Question 476
Advanced, Mixed & Conceptual Questions
The performance measure of an autonomous car could be:
  1. Number of songs played
  2. Safety, speed, comfort and fuel efficiency
  3. Number of passengers
  4. Car color preference
Correct Answer: B) Safety, speed, comfort and fuel efficiency
Question 477
Advanced, Mixed & Conceptual Questions
A reflex agent ignores percept history meaning:
  1. It uses full memory
  2. It only reacts to current input without memory
  3. It plans ahead
  4. It learns from mistakes
Correct Answer: B) It only reacts to current input without memory
Question 478
Advanced, Mixed & Conceptual Questions
A model-based agent stores:
  1. Only current percept
  2. Internal model of world state based on history
  3. Only goal state
  4. Only utility values
Correct Answer: B) Internal model of world state based on history
Question 479
Advanced, Mixed & Conceptual Questions
Goal-based agents need which capability beyond model-based agents?
  1. Sensors
  2. Actuators
  3. Search and planning to achieve goals
  4. Memory only
Correct Answer: C) Search and planning to achieve goals
Question 480
Advanced, Mixed & Conceptual Questions
Utility-based agents go beyond goal-based agents by:
  1. Not using goals
  2. Measuring desirability of states and choosing best
  3. Using simpler rules
  4. Ignoring environment
Correct Answer: B) Measuring desirability of states and choosing best
Question 481
Advanced, Mixed & Conceptual Questions
A learning agent uses a critic to:
  1. Suggest new actions
  2. Judge if performance is good or bad compared to standard
  3. Learn from textbooks
  4. Generate random actions
Correct Answer: B) Judge if performance is good or bad compared to standard
Question 482
Advanced, Mixed & Conceptual Questions
What does the performance element in a learning agent do?
  1. Learns new behaviors
  2. Makes actual decisions and actions
  3. Evaluates performance
  4. Generates training data
Correct Answer: B) Makes actual decisions and actions
Question 483
Advanced, Mixed & Conceptual Questions
What does the learning element in a learning agent do?
  1. Makes decisions
  2. Improves performance element based on critic feedback
  3. Evaluates performance
  4. Explores randomly
Correct Answer: B) Improves performance element based on critic feedback
Question 484
Advanced, Mixed & Conceptual Questions
Which component of learning agent generates new situations to learn from?
  1. Critic
  2. Performance element
  3. Learning element
  4. Problem generator
Correct Answer: D) Problem generator
Question 485
Advanced, Mixed & Conceptual Questions
In PEAS, what does Sensors describe?
  1. What agent can do
  2. How performance is measured
  3. What agent can perceive from environment
  4. Where agent operates
Correct Answer: C) What agent can perceive from environment
Question 486
Advanced, Mixed & Conceptual Questions
In PEAS, what does Actuators describe?
  1. What agent perceives
  2. How performance is measured
  3. What actions agent can take
  4. Where agent operates
Correct Answer: C) What actions agent can take
Question 487
Advanced, Mixed & Conceptual Questions
In PEAS, what does Environment describe?
  1. What agent perceives
  2. The setting in which agent operates
  3. What actions agent can take
  4. How performance is measured
Correct Answer: B) The setting in which agent operates
Question 488
Advanced, Mixed & Conceptual Questions
In PEAS, what does Performance describe?
  1. Speed of computation
  2. Criteria for success of agent
  3. Physical environment
  4. Available actions
Correct Answer: B) Criteria for success of agent
Question 489
Advanced, Mixed & Conceptual Questions
For a medical diagnosis AI: sensors would be:
  1. Prescriptions written
  2. Patient symptoms and test results
  3. Treatment recommendations
  4. Hospital schedule
Correct Answer: B) Patient symptoms and test results
Question 490
Advanced, Mixed & Conceptual Questions
For a medical diagnosis AI: actuators would be:
  1. Lab equipment
  2. Patient symptoms
  3. Diagnostic suggestions and treatment recommendations
  4. Hospital database
Correct Answer: C) Diagnostic suggestions and treatment recommendations
Question 491
Advanced, Mixed & Conceptual Questions
Simple reflex agents work well in:
  1. Complex partially observable environments
  2. Simple fully observable environments
  3. Dynamic changing environments
  4. Environments requiring planning
Correct Answer: B) Simple fully observable environments
Question 492
Advanced, Mixed & Conceptual Questions
Which agent type maintains a world model to handle partial observability?
  1. Simple reflex agent
  2. Model-based agent
  3. Learning agent only
  4. Goal-based agent only
Correct Answer: B) Model-based agent
Question 493
Advanced, Mixed & Conceptual Questions
Which AI search property means: finds a solution if one exists?
  1. Optimality
  2. Completeness
  3. Efficiency
  4. Admissibility
Correct Answer: B) Completeness
Question 494
Advanced, Mixed & Conceptual Questions
Which AI search property means: finds the BEST solution?
  1. Completeness
  2. Efficiency
  3. Optimality
  4. Admissibility
Correct Answer: C) Optimality
Question 495
Advanced, Mixed & Conceptual Questions
In which year was Artificial Intelligence officially named as a field?
  1. 1950
  2. 1956
  3. 1960
  4. 1970
Correct Answer: B) 1956
Question 496
Advanced, Mixed & Conceptual Questions
Who is considered one of the founding fathers of AI?
  1. Albert Einstein
  2. John McCarthy
  3. Isaac Newton
  4. Charles Darwin
Correct Answer: B) John McCarthy
Question 497
Advanced, Mixed & Conceptual Questions
What does the intelligent agent sense from its environment?
  1. Actuators
  2. Percepts
  3. Goals
  4. Utilities
Correct Answer: B) Percepts
Question 498
Advanced, Mixed & Conceptual Questions
Which type of AI exists today (narrow or general)?
  1. General AI (AGI)
  2. Superintelligent AI
  3. Narrow AI for specific tasks
  4. Conscious AI
Correct Answer: C) Narrow AI for specific tasks
Question 499
Advanced, Mixed & Conceptual Questions
The future of AI is expected to include:
  1. Replacing all human work immediately
  2. Explainable AI, human-AI collaboration, smart cities
  3. Stopping all development
  4. AI becoming human
Correct Answer: B) Explainable AI, human-AI collaboration, smart cities
Question 500
Advanced, Mixed & Conceptual Questions
Artificial Intelligence ultimately aims to:
  1. Replace human creativity entirely
  2. Build intelligent systems solving real-world problems and enhancing human capabilities
  3. Operate only in labs
  4. Work only with robots
Correct Answer: B) Build intelligent systems solving real-world problems and enhancing human capabilities
Converted from PDF into a WordPress-friendly HTML file with clickable hidden answers.
Share with love!

Related Posts

Leave a Reply

Your email address will not be published. Required fields are marked *