QCM : Artificial Intelligence Question Bank — 21 questions

Questions et réponses du QCM

1. How does an intelligent agent interact with its environment?

It perceives through actuators and acts through sensors
It evaluates performance through sensors and plans through actuators
It perceives through sensors and acts through actuators
It stores percepts in sensors and evaluates goals through actuators

It perceives through sensors and acts through actuators

Explication

An intelligent agent receives information about its environment through sensors and affects the environment through actuators. Sensors provide percepts, whereas actuators perform actions.

2. Which set of components is represented by the PEAS framework?

Performance measure, Environment, Actuators, and Sensors
Performance measure, Exploration, Actions, and Search
Plans, Environment, Actuators, and Solutions
Percepts, Evaluation, Actions, and States

Performance measure, Environment, Actuators, and Sensors

Explication

PEAS stands for Performance measure, Environment, Actuators, and Sensors. The performance measure evaluates the agent, while sensors provide its percepts.

3. What criterion does a rational agent use when selecting an action?

The action that requires the fewest actuators regardless of its outcome
The action expected to maximize performance given percepts and knowledge
The action that was most recently successful in a similar environment
The action that minimizes the number of percepts received by the agent

The action expected to maximize performance given percepts and knowledge

Explication

A rational agent chooses the action expected to maximize its performance measure based on its percept sequence and available knowledge.

4. What distinguishes informed search from uninformed search?

Uninformed search can operate only when every action has the same cost
Informed search expands nodes without using any problem-specific information
Informed search uses additional heuristic knowledge to guide exploration
Uninformed search uses a heuristic while informed search ignores extra knowledge

Informed search uses additional heuristic knowledge to guide exploration

Explication

Uninformed search has no problem-specific heuristic guidance, whereas informed search uses additional knowledge to direct exploration.

5. A search procedure always expands the shallowest unexpanded node before considering deeper nodes. Which procedure is this?

Greedy Best-First Search
Breadth-First Search
Depth-First Search
Uniform Cost Search

Breadth-First Search

Explication

Breadth-First Search expands nodes in order of depth, processing the shallowest unexpanded nodes before deeper ones.

6. What does Depth-First Search do when the current branch cannot lead to a solution?

It backtracks and expands another available branch
It adds a heuristic estimate to the current branch cost
It returns to the root and expands every shallow node first
It selects the frontier node with the lowest path cost

It backtracks and expands another available branch

Explication

Depth-First Search follows the deepest unexpanded node and backtracks when the current branch cannot produce a solution.

7. Which frontier node is expanded next by Uniform Cost Search?

The node with the lowest estimated remaining cost
The node at the shallowest depth
The node on the deepest available branch
The node with the lowest accumulated path cost

The node with the lowest accumulated path cost

Explication

Uniform Cost Search prioritizes the frontier node whose accumulated path cost is lowest. This differs from depth-based or heuristic-only selection.

8. What does a heuristic function estimate in a search problem?

The cost already accumulated along the path to a state
The number of actions previously executed by the agent
The probability that the current state was generated correctly
The cost of reaching a goal from a given state

The cost of reaching a goal from a given state

Explication

A heuristic estimates the remaining cost from a given state to a goal and guides heuristic search. Path cost, by contrast, records cost already incurred.

9. How does A* prioritize nodes during search?

By using only the estimated remaining cost
By combining accumulated path cost with estimated remaining cost
By selecting the deepest node and backtracking after failure
By expanding the shallowest node regardless of its costs

By combining accumulated path cost with estimated remaining cost

Explication

A* combines the cost accumulated to reach a state with a heuristic estimate of the remaining cost. Greedy best-first search uses only the heuristic estimate.

10. What is the basic operation performed repeatedly by hill climbing?

Moving from the current state to a better neighboring state
Searching the entire state space before making any move
Accepting every neighboring state regardless of its quality
Expanding the frontier node with the lowest accumulated path cost

Moving from the current state to a better neighboring state

Explication

Hill climbing repeatedly moves to a neighboring state that is better according to the optimization objective.

11. Which move does the minimax strategy select in a two-player game?

The move with the highest outcome assuming the opponent chooses randomly
The move requiring the fewest searches regardless of the opponent’s response
The move with the highest outcome assuming the opponent responds optimally against the agent
The move that maximizes the agent’s immediate score without considering later turns

The move with the highest outcome assuming the opponent responds optimally against the agent

Explication

Minimax chooses the move that maximizes the best outcome against an opponent assumed to minimize the agent’s outcome. Unlike single-agent search, it explicitly models an opposing player’s adverse choice.

12. What is the purpose of alpha-beta pruning in game-tree search?

To assign heuristic values to every possible game state
To remove branches that cannot change the final minimax decision
To guarantee that the first move examined is the optimal move
To replace the opponent’s strategy with a random selection process

To remove branches that cannot change the final minimax decision

Explication

Alpha-beta pruning eliminates game-tree branches that cannot affect the final minimax decision, reducing the number of evaluated branches without changing the selected move.

13. Which three components define a constraint satisfaction problem?

States, actions, and rewards for reaching a goal
Variables, possible domains, and constraints on value combinations
Nodes, edges, and path costs in a search graph
Facts, rules, and conclusions derived by inference

Variables, possible domains, and constraints on value combinations

Explication

A CSP consists of variables, the domains of possible values for those variables, and constraints restricting which combinations are allowed.

14. A backtracking solver assigns values to variables and then encounters a violated constraint. What should it do next?

Keep the assignment and ignore the violated constraint
Return to an earlier assignment and try a different value
Restart the entire search without retaining any assignments
Delete every value from the remaining variable domains

Return to an earlier assignment and try a different value

Explication

Backtracking assigns variables incrementally and returns to an earlier assignment when a constraint is violated, allowing an alternative value to be tried.

15. What distinguishes a knowledge-based agent from a simple reflex agent?

It selects actions solely from the current percept
It avoids representing information about the environment
It uses stored facts and rules with inference to choose actions
It chooses actions randomly when several options are available

It uses stored facts and rules with inference to choose actions

Explication

A knowledge-based agent stores representations of facts and rules and reasons over them to select actions. A simple reflex agent relies only on the current percept.

16. Which statement best describes propositional logic?

It represents whole propositions as true or false and combines them with connectives
It represents uncertainty by assigning probabilities to hypotheses
It represents objects and relations using variables and quantified statements
It represents actions as paths through a game tree

It represents whole propositions as true or false and combines them with connectives

Explication

Propositional logic represents propositions that are either true or false and combines them using logical connectives. Objects, relations, and quantified statements are features of First-Order Logic.

17. Which capability is characteristic of First-Order Logic?

Representing only complete propositions with no internal structure
Representing objects, relations, functions, predicates, and quantified statements
Applying game strategies against an optimizing opponent
Assigning probability values to every possible conclusion

Representing objects, relations, functions, predicates, and quantified statements

Explication

First-Order Logic can describe objects, predicates, relations, functions, and quantified statements within a domain, providing more expressive structure than propositional logic.

18. How does forward chaining derive a new conclusion?

It repeatedly applies rules whose premises match known facts
It selects a move by assuming an opponent minimizes the outcome
It assigns probabilities to competing hypotheses using prior beliefs
It begins with a goal and searches backward for supporting rules

It repeatedly applies rules whose premises match known facts

Explication

Forward chaining starts with known facts and repeatedly applies rules whose premises match those facts, producing new conclusions.

19. When does uncertainty arise in artificial intelligence?

When an agent has no rules for representing knowledge at all
When an agent cannot determine the true state or outcome with complete confidence
When every proposition has already been proven true or false
When known conditions always produce a fixed and predictable outcome

When an agent cannot determine the true state or outcome with complete confidence

Explication

Uncertainty occurs when the agent lacks complete confidence about the true state or outcome. Deterministic reasoning instead assumes fixed outcomes from known conditions.

20. What does Bayesian inference do with evidence and prior beliefs?

It assigns conclusions without representing degrees of belief
It updates the probabilities of hypotheses
It eliminates the need to represent alternative hypotheses
It guarantees that every hypothesis receives the same probability

It updates the probabilities of hypotheses

Explication

Bayesian inference updates the probability of hypotheses by combining new evidence with prior beliefs, thereby representing degrees of belief rather than deterministic conclusions.

21. What does a Bayesian network represent?

Logical contradictions among propositions in an undirected structure
Exact assignments of values to variables in a constraint problem
The sequence of moves selected by two competing game players
Probabilistic dependencies among variables in a directed graphical structure

Probabilistic dependencies among variables in a directed graphical structure

Explication

A Bayesian network uses a directed graphical structure to represent probabilistic dependencies among variables.

Révisez avec les flashcards

Mémorisez les réponses avec 39 flashcards sur Artificial Intelligence Question Bank.

How does an intelligent agent perceive its environment?

Through sensors.

What does the PEAS framework specify for an agent?

Performance measure, Environment, Actuators, and Sensors.

What action does a rational agent select?

The action expected to maximize its performance measure.

Voir les flashcards →

Approfondir avec la fiche

Consultez la fiche de révision complète sur Artificial Intelligence Question Bank.

Voir la fiche →

Cours similaires

Introduction to Computer Programming

Programmation

Introduction to Computer Programming

Programmation

Protocoles de communication réseau

Réseaux

Fondamentaux de FastAPI

Programmation

UX Design et conception centrée utilisateur

Python Basics and Math Functions

Programmation

Crée tes propres QCM

Importe ton cours et l'IA génère des QCM avec corrections en 30 secondes.

Générateur de QCM