WorksheetsPrelim Exam - A.I. - Part 1
Total questions: 40
Worksheet time: 30mins
Which of the following best defines a search space?
The final solution of a problem
A set of possible solutions a system may have
The memory required for searching
The sequence of percepts an agent receives
The start state in a search problem refers to:
The most optimal solution
The initial point where the agent begins searching
The goal state of the problem
The path cost of the first step
Which property of a search algorithm guarantees finding a solution if one exists?
Optimality
Completeness
Time complexity
Space complexity
The function that checks if the current state meets the goal is called:
Start state
Search tree
Goal test
Transition model
In a search tree, the “fringe” refers to:
The root node
All expanded nodes
The leaves (nodes available for expansion)
The internal states already visited
Which search strategy explores all nodes at the current depth before moving deeper?
Depth-First Search
Breadth-First Search
Hill Climbing
A* Search
A search algorithm is considered optimal if it:
Finds the fastest solution
Always finds a solution
Guarantees the best (lowest cost) solution
Uses the least amount of memory
Heuristic functions are used in search to:
Guarantee completeness
Guide search using problem-specific knowledge
Increase space complexity
Replace the goal test
Which uninformed search strategy combines DFS space efficiency with BFS completeness?
Greedy Best-First Search
Iterative Deepening DFS
Hill Climbing
A* Search
The A* algorithm combines the benefits of:
Depth-First Search and Hill Climbing
Greedy Best-First and Uniform-Cost Search
Breadth-First Search and Hill Climbing
Heuristics and Random Search
Which is NOT a factor when choosing a search algorithm?
Completeness
Optimality
Reproducibility
Time complexity
Which of the following is an informed search algorithm?
Depth-First Search
Iterative Deepening
Greedy Best-First Search
Breadth-First Search
Heuristic algorithms are especially useful in:
Small, finite problems
Problems with guaranteed solutions
Complex real-world problems with limited time/space
Problems where the goal is unknown
The main disadvantage of uninformed (blind) search is:
Lack of completeness
Excessive time and space use
Dependence on heuristics
Failure to guarantee a goal
The set of valid states for a problem is called:
Search tree
State space
Transition model
Goal function
An agent is defined as:
Hardware only
A set of rules for solving problems
Architecture + Program
Percepts + Actions
The program of an agent is responsible for:
Producing percepts
Mapping percepts to actions
Controlling sensors
Designing architecture
Which type of agent reacts only to the current percept?
Model-based reflex agent
Simple reflex agent
Goal-based agent
Utility-based agent
A model-based reflex agent differs from a simple reflex agent because it:
Ignores percept history
Operates only in fully observable environments
Maintains an internal state based on percept history
Maximizes happiness among goals
Which type of agent is designed to maximize “happiness” or utility?
Reflex agent
Goal-based agent
Utility-based agent
Model-based reflex agent
A problem-solving agent typically goes through which sequence?
Execution → Problem formulation → Goal formulation → Search
Goal formulation → Problem formulation → Search → Execution
Perception → Action → Execution → Goal test
Transition model → Goal test → Path cost → Execution
Which of the following is NOT part of a well-defined problem?
Initial state
Path cost
Architecture
Transition model
In the 8-puzzle problem, the path cost usually refers to:
The number of tiles in the puzzle
The total number of states explored
The number of steps taken to reach the goal
The optimal heuristic used
A goal-based agent makes decisions by:
Acting solely on percepts
Tracking internal state only
Selecting actions that achieve a predefined goal
Maximizing reward among alternatives
Utility-based agents are most beneficial when:
There is only one possible goal
Multiple alternatives exist and trade-offs must be evaluated
The environment is fully observable
Path cost is not measurable
A transition model in problem-solving defines:
The path cost
What each action does and the resulting state
The sequence of goals
The heuristic function
Which agent type is most suitable for partially observable environments?
Simple reflex agent
Model-based reflex agent
Goal-based agent
Utility-based agent
Which of the following is a common AI application in medicine?
Assembly line automation
Patient monitoring and diagnosis
Fraud detection
Personalized learning
The PEAS framework describes:
The cycle of problem-solving agents
The task environment of an agent
The rules of heuristic search
The laws of thought in AI
In the PEAS framework, “Sensors” correspond to:
Decision-making tools
Perception of the environment
The hardware architecture
The path cost function
Which environment type is unpredictable in its outcomes?
Deterministic
Stochastic
Episodic
Static
The Turing Test evaluates AI based on:
Logical reasoning ability
Human-like observable behavior
Rational action outcomes
Internal computational models
Which AI perspective focuses on simulating human thought processes?
Rational Agent
Laws of Thought
Cognitive Science
Turing Test
A rational agent differs from an omniscient agent because it:
Knows the future
Acts optimally with available information
Always mimics humans
Relies only on pre-programmed rules
An example of a software agent’s sensors includes:
Keystrokes and network packets
Cameras and range finders
Eyes and ears
Motors and actuators
Which component of AI systems involves inference and decision-making?
Perception
Reasoning
Action
Transition model
Autonomy in agents means:
Relying entirely on the designer’s initial knowledge
Depending only on sensors without memory
Behavior is determined significantly by its own experience
Performing actions without reasoning
Which environment type involves independent episodes, simplifying decision-making?
Dynamic
Non-episodic
Episodic
Continuous
In AI applications, which industry uses AI for fraud detection?
Education
Medicine
Business & Finance
Engineering
The performance measure of a rational agent is used to:
Define the agent’s architecture
Evaluate how successful its actions are
Map percepts to actions
Replace the environment model
