What is one of the primary goals of Artificial Intelligence?

A) To create human-like robots

B) To make computers more useful by automating dangerous tasks

C) To replace human jobs entirely

D) To develop advanced video games

Correct Answer: B

Who is considered the founding father of AI along with Alan Turing?

A) McCarthy

B) Minsky

C) McCulloch

D) Zadeh

Correct Answer: C

What event in 1956 is often considered the birth of Artificial Intelligence?

A) The publication of Turing's paper

B) The Dartmouth meeting

C) The introduction of the Lisp language

D) The first AI winter

Correct Answer: B

Which of the following is an application of AI?

A) Manual data entry

B) Speech recognition

C) Simple calculators

D) Traditional bookkeeping

Correct Answer: B
What concept did Lotfi A. Zadeh publish in 1965 that is fundamental to
AI?

A) Neural networks

B) Quantum computing

C) Fuzzy sets

D) Expert systems

Correct Answer: C

True/False Questions

The definition of intelligence includes the ability to understand and learn

things.

Correct Answer: True

AI applications include only robotics and face recognition.

Correct Answer: False

The first expert system, DENDRAL, was developed to determine

molecular structures.

Correct Answer: True

The term "Artificial Intelligence" was first coined in the late 1980s.

Correct Answer: False

Turing's Imitation Game is a test to evaluate a machine's ability to

exhibit intelligent behavior.

Correct Answer: True

Which discipline does NOT contribute to the foundation of Artificial

Intelligence?

A) Philosophy

B) Biology

C) Mathematics

D) Linguistics

Correct Answer: B
What technology was introduced by Hopfield in 1982?

A) Fuzzy logic

B) Expert systems

C) Neural networks with feedback

D) Machine learning

Correct Answer: C

Which AI application involves the identification and extraction of

information from text?

A) Image processing

B) Natural language processing

C) Expert systems

D) Robotics

Correct Answer: B

What significant change in AI occurred in the mid-1980s?

A) The first AI winter

B) The rise of expert systems

C) The rebirth of neural networks

D) The invention of Lisp

Correct Answer: C

Which of the following applications uses AI for recognizing human faces?

A) Speech recognition

B) Computer vision

C) Text classification

D) Fuzzy logic

Correct Answer: B

True/False Questions
 The AI winter refers to a period of increased funding and enthusiasm for
artificial intelligence research.

Correct Answer: False

Alan Turing's work laid the groundwork for modern computational

theories, including AI.

Correct Answer: True

MYCIN was an early expert system designed for medical diagnosis.

Correct Answer: True

Fuzzy logic is unrelated to artificial intelligence applications.

Correct Answer: False

Machine learning and natural language processing became prominent

fields in the 1990s within AI.

Correct Answer: True

What does AI primarily aim to do with tasks that are dangerous or

tedious for humans?

A) Eliminate them entirely

B) Automate them

C) Simplify them

D) Enhance them

Correct Answer: B

In which year was the first mathematical model of a neural network

published?

A) 1950

B) 1943

C) 1965

D) 1970

Correct Answer: B
Which of the following is NOT a type of AI application mentioned in the
chapter?

A) Image processing

B) Virtual reality

C) Face recognition

D) Natural language processing

Correct Answer: B

What was the main focus of the Dartmouth meeting in 1956?

A) Development of robotics

B) Establishing AI as a field of study

C) Funding for AI projects

D) Publishing research papers

Correct Answer: B

Which technology is used for converting text into spoken words?

A) Machine translation

B) Speech synthesis

C) Information extraction

D) Text classification

Correct Answer: B

True/False Questions

The concept of artificial intelligence has been studied for less than 100
years.

Correct Answer: False

Control theory is important for AI because it involves feedback from the

environment.

Correct Answer: True

 The British government was fully supportive of AI research throughout
the 1960s.

Correct Answer: False

Genetic algorithms were introduced after the first AI winter.

Correct Answer: False

Speech recognition technology is a part of natural language processing

applications.

Correct Answer: True

Which of the following best defines Artificial Intelligence according to

Alan Turing?

A) The study of human behavior

B) The science and engineering of making intelligent machines

C) The development of advanced robotics

D) The programming of computers to play games

Correct Answer: B

What is a significant effect of the first AI winter?

A) Increased funding for AI research

B) Suspension of government support for AI projects

C) Development of new AI technologies

D) Rise in public interest in AI

Correct Answer: B

Which of the following systems was designed for mineral exploration?

A) MYCIN

B) PROSPECTOR

C) DENDRAL

D) Hopfield network

Correct Answer: B
What major advancement in AI occurred in the 2010s?

A) Introduction of expert systems

B) Rise of massive data and new computing power

C) Development of neural networks

D) Creation of the Lisp language

Correct Answer: B

Which field studies the phenomena of perception relevant to AI?

A) Economics

B) Linguistics

C) Psychology

D) Control theory

Correct Answer: C

True/False Questions

Fuzzy logic was developed to handle uncertain or imprecise information.

Correct Answer: True

The introduction of the Lisp programming language occurred in the

early 2000s.

Correct Answer: False

The goals of AI include designing programs that can solve problems like
humans.

Correct Answer: True

Neural networks became popular again in the mid-1980s after a period of

decline.

Correct Answer: True

The concept of a fuzzy set has no practical applications in AI.

Correct Answer: False

What is an agent in the context of AI?

A) A computer program without sensors

B) Anything that perceives its environment through sensors and acts upon it
through actuators

C) A type of neural network

D) A mathematical model of decision-making

Correct Answer: B

Which of the following best describes a rational agent?

A) An agent that knows everything

B) An agent that maximizes expected performance based on its percepts

C) An agent that always makes perfect decisions

D) An agent that performs actions randomly

Correct Answer: B

What does the acronym PEAS stand for?

A) Performance, Environment, Actions, Sensors

B) Perception, Energy, Actions, Sensors

C) Performance measure, Environment, Actuators, Sensors

D) Perceptual, Environmental, Actuator, Sensor

Correct Answer: C

Which type of agent uses a model of the world to choose actions based on
internal states?

A) Simple reflex agent

B) Model-based reflex agent

C) Goal-based agent

D) Utility-based agent

Correct Answer: B

What is the primary function of a learning agent?

A) To perform actions automatically

B) To improve its performance over time using feedback

C) To maintain constant behavior regardless of the environment

D) To gather data without acting

Correct Answer: B

True/False Questions

An agent’s percept sequence is the complete history of everything the

agent has perceived.

Correct Answer: True

A simple reflex agent considers the entire history of percepts when

making decisions.

Correct Answer: False

The autonomy of an agent is determined by its reliance on the designer's

knowledge.

Correct Answer: False

Goal-based agents track their goals and select actions that lead to
achieving those goals.

Correct Answer: True

A utility-based agent chooses actions that maximize its expected utility

based on a defined utility function.

Correct Answer: True

What type of environment does a mobile robot typically operate in?

A) Static and unchanging

B) Dynamic with obstacles

C) Fully observable only

D) Only indoor spaces

Correct Answer: B

What is the primary goal of a vacuum-cleaner agent in its environment?

A) To navigate to different locations

B) To clean dirty areas

C) To avoid all obstacles

D) To analyze data

Correct Answer: B

Which agent is characterized by its ability to learn from past experiences?

A) Simple reflex agent

B) Model-based reflex agent

C) Learning agent

D) Goal-based agent

Correct Answer: C

In the PEAS framework, what describes the “environment” component?

A) The actions the agent can perform

B) The agent's past experiences

C) The context in which the agent operates

D) The performance measure of success

Correct Answer: C

Which of the following properties describes a fully observable

environment?

A) The agent cannot access complete information about the environment

B) The agent can perceive all aspects of the environment at any time

C) The environment changes unpredictably

D) The agent operates in isolation without feedback

Correct Answer: B

True/False Questions

A rational agent always knows the actual outcome of its actions.

Correct Answer: False

Reflex agents do not maintain any internal state or memory.

Correct Answer: True

The performance measure evaluates how well an agent is performing

based on environmental states.

Correct Answer: True

In a stochastic environment, the outcomes of actions are predictable and

certain.

Correct Answer: False

Utility-based agents make decisions based solely on immediate rewards.

Correct Answer: False

What characterizes a goal-based agent?

A) It acts solely based on reflexes.

B) It maintains a set of goals and chooses actions to achieve them.

C) It operates without any internal model.

D) It does not learn from past experiences.

Correct Answer: B

Which component of the PEAS framework refers to the actions an agent

can perform?

A) Performance measure

B) Environment

C) Actuators

D) Sensors

Correct Answer: C

What distinguishes a utility-based agent from a goal-based agent?

A) It has no goals.

B) It uses a utility function to measure preferences among states.

C) It only acts on immediate percepts.

D) It does not learn from experiences.

Correct Answer: B

In which type of environment would a vacuum-cleaner agent most likely

operate?

A) Fully observable and static

B) Partially observable and dynamic

C) Completely predictable

D) Unchanging and isolated

Correct Answer: B

What is the role of the critic in a learning agent?

A) To execute actions

B) To provide feedback on the agent's performance

C) To define the agent's goals

D) To gather data

Correct Answer: B

True/False Questions

An agent function maps each percept sequence to a specific action.

Correct Answer: True

A model-based reflex agent ignores the current state of the world.

Correct Answer: False

Learning agents can improve their performance based on feedback from

their environment.

Correct Answer: True

The autonomy of an agent is determined by its reliance on external

information.

Correct Answer: False

In a discrete environment, actions and states are limited and countable.

Correct Answer: True

What is the primary function of an actuator in an intelligent agent?

A) To perceive the environment

B) To process information

C) To perform actions in the environment

D) To store data

Correct Answer: C

Which term refers to the complete history of everything an agent has ever
perceived?

A) Percept

B) Percept sequence

C) Action history

D) Environment state

Correct Answer: B

What is a characteristic of a simple reflex agent?

A) It updates its internal model continuously.

B) It selects actions based only on the current percept.

C) It is capable of learning from experiences.

D) It maintains a long-term memory of past actions.

Correct Answer: B

Which of the following environments is considered dynamic?

A) A classroom

B) A busy street with moving vehicles

C) A library

D) A computer program running in isolation

Correct Answer: B

What is the purpose of a utility function in a utility-based agent?

A) To evaluate the environment’s current state

B) To determine the best action based on expected outcomes

C) To store perceptual data

D) To maintain a list of goals

Correct Answer: B

True/False Questions

A rational agent always performs the best action based on its knowledge
of the environment.

Correct Answer: True

The environment in which an agent operates can influence its decision-

making process.

Correct Answer: True

All agents are required to have sensors and actuators to be considered

intelligent.

Correct Answer: True

A learning agent is unable to adapt its behavior based on previous

experiences.

Correct Answer: False

The concept of rationality in AI includes making decisions that maximize

the expected performance measure.

Correct Answer: True

What is the primary role of search algorithms in AI?

A) To create neural networks

B) To solve problems by exploring different options

C) To enhance visual recognition

D) To store data efficiently

Correct Answer: B

What does the term "search space" refer to in a search problem?

A) The starting point of the search

B) The set of all possible solutions

C) The final goal of the search

D) The algorithm used to find the solution

Correct Answer: B

Which of the following search strategies is known as "blind search"?

A) Informed search

B) Heuristic search

C) Uninformed search

D) Optimal search

Correct Answer: C

What is a key characteristic of the A search algorithm?*

A) It only uses depth-first search

B) It combines path cost and heuristic values

C) It guarantees the shortest path without any conditions

D) It is always faster than breadth-first search

Correct Answer: B

Which search algorithm guarantees finding a solution if one exists?

A) Depth-First Search (DFS)

B) Breadth-First Search (BFS)

C) Uniform-Cost Search (UCS)

D) All of the above

Correct Answer: D
True/False Questions

Depth-First Search (DFS) uses a first-in, first-out (FIFO) approach to

enqueue nodes.

Correct Answer: False

A search algorithm is considered complete if it guarantees finding a

solution whenever one exists.

Correct Answer: True

Beam Search is guaranteed to find the optimal solution.

Correct Answer: False

The performance of search algorithms can be evaluated based on

completeness, time complexity, and space complexity.

Correct Answer: True

Informed search strategies do not utilize any information about the likely
direction of the goal.

Correct Answer: False

What type of search strategy is Breadth-First Search (BFS)?

A) Informed search

B) Depth-first search

C) Uninformed search

D) Heuristic search

Correct Answer: C

Which search algorithm is optimal when all operators have the same
cost?

A) Depth-First Search (DFS)

B) Uniform-Cost Search (UCS)

C) A* Search

D) Greedy Search

Correct Answer: B
What is one of the main advantages of Depth-First Search (DFS)?

A) Guarantees finding the shortest path

B) Requires very little memory

C) Always finds the optimal solution

D) Explores all nodes at the same level first

Correct Answer: B

In which scenario would you use Depth-First Iterative Deepening

(DFID)?

A) When memory usage is not a concern

B) When the maximum depth of the solution is unknown

C) When you need to find the shortest path quickly

D) When only a single solution is required

Correct Answer: B

What is the purpose of a heuristic function in informed search

algorithms?

A) To ensure completeness

B) To estimate the cost from a node to the goal

C) To increase memory usage

D) To create the search space

Correct Answer: B

True/False Questions

Uniform-Cost Search (UCS) focuses solely on the number of nodes

expanded.

Correct Answer: False

A* Search is both optimal and complete if the heuristic is admissible.

Correct Answer: True

Beam Search evaluates all nodes at each level of the search tree.
Correct Answer: False

Depth-Limited Search is a variant of Depth-First Search that imposes a

maximum depth limit.

Correct Answer: True

All informed search strategies are guaranteed to find a solution.

Correct Answer: False

What is the main disadvantage of Breadth-First Search (BFS)?

A) Requires little memory

B) Can find long solutions quickly

C) Requires a large amount of memory

D) Always finds the optimal solution

Correct Answer: C

Which search algorithm can enter an infinite loop if not properly

bounded?

A) Uniform-Cost Search (UCS)

B) Depth-First Search (DFS)

C) Breadth-First Search (BFS)

D) A* Search

Correct Answer: B

What does the term “path cost” refer to in the context of search
algorithms?

A) The total number of nodes expanded

B) The cumulative cost of reaching a node from the start

C) The heuristic estimate from the node to the goal

D) The maximum depth of the search tree

Correct Answer: B

In Beam Search, what determines which nodes are expanded?

A) The depth of the nodes

B) The number of nodes in the search space

C) The heuristic values of the nodes

D) The path cost of nodes

Correct Answer: C

Which search algorithm is characterized by using a priority queue?

A) Depth-First Search (DFS)

B) Breadth-First Search (BFS)

C) Uniform-Cost Search (UCS)

D) All of the above

Correct Answer: C

True/False Questions

Uniform-Cost Search (UCS) is complete and optimal for all types of

graphs.

Correct Answer: True

An informed search strategy is guaranteed to find the shortest path to the

goal.

Correct Answer: False

The search tree's root node corresponds to the initial state of the
problem.

Correct Answer: True

Depth-First Iterative Deepening (DFID) combines the advantages of both

BFS and DFS.

Correct Answer: True

A heuristic function can overestimate the true cost to reach the goal in A*
Search.

Correct Answer: False

What is a key feature of the A search algorithm?*

A) It uses a fixed path cost only

B) It combines path cost and heuristic estimates

C) It performs depth-first search exclusively

D) It is not optimal under any conditions

Correct Answer: B

Which search strategy is most suitable for large state spaces where the
solution depth is unknown?

A) Breadth-First Search (BFS)

B) Depth-First Search (DFS)

C) Depth-First Iterative Deepening (DFID)

D) Uniform-Cost Search (UCS)

Correct Answer: C

In which search algorithm does the agent explore all possible nodes at the
current depth before moving deeper?

A) Depth-First Search (DFS)

B) Uniform-Cost Search (UCS)

C) Breadth-First Search (BFS)

D) A* Search

Correct Answer: C

What is the main purpose of the “goal test” in a search algorithm?

A) To define the search space

B) To evaluate the efficiency of the algorithm

C) To determine if the current state is the goal state

D) To assign costs to actions

Correct Answer: C

Which of the following statements about Depth-First Search (DFS) is

true?
A) It guarantees finding the shortest path.

B) It requires more memory than BFS.

C) It can be faster than BFS in certain scenarios.

D) It is always complete.

Correct Answer: C

True/False Questions

A search algorithm is considered optimal if it always finds the best

possible solution.

Correct Answer: True

Informed search algorithms use heuristics to direct their search towards

the goal.

Correct Answer: True

Beam Search is guaranteed to find the optimal solution every time.

Correct Answer: False

The transition model describes the actions available to an agent in a

search problem.

Correct Answer: True

All search algorithms have the same time and space complexity
regardless of their approach.

Correct Answer: False

What is a characteristic of rule-based expert systems?

A) They use neural networks for learning.

B) They operate solely on fuzzy logic.

C) They utilize "if-then" rules for decision-making.

D) They do not require human expertise.

Correct Answer: C

Which expert system is known for diagnosing bacterial infections?

A) Fuzzy Logic Systems

B) MYCIN

C) Neural Network-Based Systems

D) Frame-Based Systems

Correct Answer: B

What does the knowledge base in a rule-based expert system contain?

A) User interface elements

B) A set of facts and data

C) The rules for problem-solving

D) An explanation facility

Correct Answer: C

What is the role of a knowledge engineer in an expert system development

team?

A) To develop the user interface

B) To write the rules and algorithms

C) To design, build, and test the expert system

D) To provide domain expertise

Correct Answer: C

Which of the following is a benefit of expert systems?

A) They can replace all human experts.

B) They provide inconsistent recommendations.

C) They are available 24/7 for support.

D) They require no maintenance once developed.

Correct Answer: C

True/False Questions

Expert systems can only operate in well-defined and narrow domains.

Correct Answer: True

The end-user is responsible for developing the expert system.

Correct Answer: False

Backward chaining starts with a hypothesis and works backward to find

supporting facts.

Correct Answer: True

Expert systems can make mistakes, similar to human experts.

Correct Answer: True

Fuzzy logic systems provide binary true/false outputs.

Correct Answer: False

What type of expert system uses frames for knowledge representation?

A) Rule-Based Expert Systems

B) Frame-Based Expert Systems

C) Fuzzy Logic Systems

D) Neural Network-Based Expert Systems

Correct Answer: B

What is the primary function of the inference engine in an expert system?

A) To store user data

B) To represent knowledge

C) To carry out reasoning and reach conclusions

D) To provide technical support

Correct Answer: C

What is a neuro-fuzzy expert system?

A) A system that uses only fuzzy logic

B) A combination of neural networks and fuzzy logic

C) A traditional rule-based system

D) A system that does not make predictions

Correct Answer: B

Which component of an expert system enables users to understand how

conclusions are reached?

A) Knowledge base

B) Inference engine

C) Explanation facilities

D) User interface

Correct Answer: C

In which application area are expert systems commonly used for

diagnosing medical conditions?

A) Manufacturing

B) Financial Services

C) Medical Diagnosis

D) Technical Support

Correct Answer: C

True/False Questions

The project manager is responsible for the technical aspects of coding in

an expert system.

Correct Answer: False

Expert systems can provide consistent and reliable recommendations.

Correct Answer: True

Forward chaining is a goal-driven approach to problem-solving.

Correct Answer: False

The knowledge limitation of expert systems can affect their performance.

Correct Answer: True

Maintenance of expert systems is not necessary once they are developed.

Correct Answer: False

What is the main purpose of rules in a rule-based expert system?

A) To store user preferences

B) To provide a structured way to solve problems

C) To enhance computational speed

D) To predict future trends

Correct Answer: B

Which of the following best describes fuzzy logic systems?

A) They operate on binary values only.

B) They handle precise and certain information.

C) They manage uncertain or imprecise information.

D) They are only used in manufacturing.

Correct Answer: C

Who is primarily responsible for ensuring that the expert system meets
user needs?

A) Domain expert

B) Knowledge engineer

C) Project manager

D) End-user

Correct Answer: D

What is the significance of knowledge preservation in expert systems?

A) It allows systems to function without updates.

B) It ensures that expertise is retained and accessible.

C) It eliminates the need for human experts.

D) It reduces the complexity of the system.

Correct Answer: B
Which method is typically used for diagnosing issues in expert systems?

A) Forward Chaining

B) Backward Chaining

C) Frame Representation

D) Fuzzy Logic

Correct Answer: B

True/False Questions

An expert system shell requires users to develop all components from

scratch.

Correct Answer: False

Expert systems are designed to operate in broad and undefined domains.

Correct Answer: False

The user interface is critical for the successful adoption of an expert

system.

Correct Answer: True

Expert systems can adapt to new information without any updates

required.

Correct Answer: False

The knowledge engineer interviews domain experts to capture their

problem-solving methods.

Correct Answer: True

Which of the following is a key characteristic of expert systems?

A) They require constant human intervention to make decisions.

B) They operate at the same level as human experts in specific domains.

C) They are always superior to human decision-making.

D) They function without any rules or guidelines.

Correct Answer: B
What type of knowledge representation technique do expert systems
primarily use?

A) Neural networks

B) Frames

C) Production rules

D) Fuzzy sets

Correct Answer: C

In expert systems, what does the term "user interface" refer to?

A) The backend database of the system

B) The means through which users interact with the system

C) The rules that govern system behavior

D) The hardware required to run the system

Correct Answer: B

What role does the project manager play in the development of an expert
system?

A) They create the knowledge base.

B) They lead the development team and ensure project milestones are met.

C) They encode the rules into the system.

D) They act as the domain expert.

Correct Answer: B

Which statement best describes the limitations of expert systems?

A) They can operate in any environment without limitations.

B) They are immune to errors and always provide accurate output.

C) Their effectiveness depends on the accuracy and completeness of the

knowledge base.

D) They automatically update their knowledge base.

Correct Answer: C
True/False Questions

The knowledge engineer is responsible for maintaining the system after

its development.

Correct Answer: True

Expert systems can only process definite and precise data.

Correct Answer: False

Forward chaining uses existing facts to infer new conclusions.

Correct Answer: True

Expert systems do not require updates once they are fully developed.

Correct Answer: False

The end-user plays a crucial role in the design of the expert system's user
interface.

Correct Answer: True

What is the primary inspiration behind genetic algorithms?

A) Newton's laws of motion

B) Darwin's theory of evolution

C) Quantum mechanics

D) Artificial neural networks

Correct Answer: B

Who introduced genetic algorithms in the 1960s?

A) Alan Turing

B) John Holland

C) David E. Goldberg

D) Herbert Simon

Correct Answer: B

In genetic algorithms, what does the fitness function measure?

A) The speed of the algorithm

B) The quality of a solution

C) The diversity of the population

D) The complexity of the problem

Correct Answer: B

Which selection mechanism involves choosing individuals based on their

fitness scores?

A) Random Selection

B) Tournament Selection

C) Rank Selection

D) Roulette Wheel Selection

Correct Answer: D

What is the purpose of mutation in genetic algorithms?

A) To combine two parent solutions

B) To maintain diversity in the population

C) To evaluate fitness scores

D) To select the best individuals

Correct Answer: B

True/False Questions

Genetic algorithms guarantee finding the absolute best solution to a

problem.

Correct Answer: False

Crossover in genetic algorithms involves creating offspring by combining

genetic material from two parents.

Correct Answer: True

The population in a genetic algorithm consists of only one solution.

Correct Answer: False

 Genetic algorithms can be applied to both continuous and discrete
optimization problems.

Correct Answer: True

Premature convergence in genetic algorithms can lead to suboptimal

solutions.

Correct Answer: True

What kind of encoding technique is commonly used in genetic

algorithms?

A) Hexadecimal encoding

B) Binary encoding

C) ASCII encoding

D) Octal encoding

Correct Answer: B

Which type of crossover selects a single point to swap genetic material?

A) Multi-Point Crossover

B) Uniform Crossover

C) Single-Point Crossover

D) Random Crossover

Correct Answer: C

What happens during the convergence of a genetic algorithm?

A) The population becomes more diverse.

B) The individuals in the population become similar.

C) The algorithm generates new random solutions.

D) The performance of the algorithm degrades.

Correct Answer: B

What is a common application of genetic algorithms in machine learning?

A) Data cleaning
B) Evolving decision trees

C) Feature extraction

D) Clustering analysis

Correct Answer: B

What is one of the main limitations of genetic algorithms?

A) They can only solve linear problems.

B) They are computationally expensive.

C) They require no parameter tuning.

D) They always find the optimal solution.

Correct Answer: B

True/False Questions

Genetic algorithms are guaranteed to explore the entire search space.

Correct Answer: False

The mutation rate in genetic algorithms is typically set very high to

ensure diversity.

Correct Answer: False

Tournament selection involves randomly selecting a group of individuals

and choosing the best one from that group.

Correct Answer: True

Genetic algorithms can be applied in financial forecasting and portfolio

optimization.

Correct Answer: True

Inversion mutation involves shuffling the values of selected genes in a

chromosome.

Correct Answer: False

Which of the following best describes the role of the fitness function in
genetic algorithms?

A) It selects parents for reproduction.

B) It measures how well a solution solves the problem.

C) It encodes the genetic information.

D) It determines the mutation rate.

Correct Answer: B

What is the primary goal of genetic algorithms?

A) To create random solutions

B) To find approximate solutions to complex problems

C) To eliminate all errors in data

D) To guarantee optimal solutions

Correct Answer: B

Which mutation technique randomly flips bits in a binary chromosome?

A) Random Resetting

B) Bit Flip Mutation

C) Scramble Mutation

D) Inversion Mutation

Correct Answer: B

What is a potential consequence of low diversity in a genetic algorithm

population?

A) Faster convergence to a solution

B) Higher likelihood of reaching local optima

C) Improved solution quality

D) More exploration of the search space

Correct Answer: B

Which of the following is a common termination criterion for genetic

algorithms?

A) Randomly stopping after a set time

B) Reaching a predefined number of generations

C) Selecting the first solution found

D) Always improving the fitness score

Correct Answer: B

True/False Questions

Genetic algorithms require gradient information to optimize solutions.

Correct Answer: False

Population size can significantly affect the performance of a genetic

algorithm.

Correct Answer: True

Genetic algorithms can only be applied to optimization problems in

engineering.

Correct Answer: False

Roulette Wheel Selection gives higher fitness individuals a greater chance

of being selected.

Correct Answer: True

The process of crossover in genetic algorithms is similar to biological

reproduction.

Correct Answer: True

Which of the following is a characteristic of single-point crossover?

A) It selects multiple points for swapping genes.

B) It treats each gene independently.

C) It swaps genetic material at one specific point.

D) It requires no crossover at all.

Correct Answer: C

What is the function of the mutation rate in genetic algorithms?

A) To determine the size of the population

B) To control how often genetic material is combined

C) To introduce variability into the population

D) To select the fittest individuals for breeding

Correct Answer: C

Which of the following methods is NOT typically used in genetic

algorithms?

A) Gradient descent

B) Tournament selection

C) Rank selection

D) Random selection

Correct Answer: A

In a genetic algorithm, what is meant by "convergence"?

A) The algorithm stops because no solutions are found.

B) The fitness of the population becomes stagnant over generations.

C) The population becomes diverse and explores new areas.

D) The algorithm randomly selects solutions.

Correct Answer: B

Which of the following is a real-world application of genetic algorithms?

A) Image processing

B) Gene sequencing

C) Text mining

D) Data entry automation

Correct Answer: B

True/False Questions

Genetic algorithms can effectively solve both linear and nonlinear

problems.
Correct Answer: True

The fitness function is irrelevant to the selection process in genetic

algorithms.

Correct Answer: False

Genetic algorithms are not suitable for problems with large search
spaces.

Correct Answer: False

Maintaining diversity in a genetic algorithm's population helps prevent

premature convergence.

Correct Answer: True

Genetic algorithms can only be implemented in programming languages

like Python and Java.

Correct Answer: False

What is the primary focus of machine learning?

A) Writing explicit programs for tasks

B) Creating systems that learn from data

C) Performing manual data analysis

D) Developing hardware for AI

Correct Answer: B

Which type of machine learning uses labeled data for training?

A) Unsupervised Learning

B) Reinforcement Learning

C) Supervised Learning

D) Deep Learning

Correct Answer: C

What is an example of unsupervised learning?

A) Email spam detection

B) Customer segmentation

C) Stock price prediction

D) Disease diagnosis

Correct Answer: B

Which algorithm is primarily used for classification tasks in supervised

learning?

A) K-means

B) Decision Trees

C) Principal Component Analysis

D) DBSCAN

Correct Answer: B

What does the 'K' in K-nearest neighbors (KNN) refer to?

A) The number of features used

B) The number of nearest neighbors considered

C) The total number of classes

D) The size of the dataset

Correct Answer: B

True/False Questions

Machine learning can operate without explicit programming for each

specific task.

Correct Answer: True

Deep learning is a type of supervised learning.

Correct Answer: True

In supervised learning, models learn from unlabeled data.

Correct Answer: False

Data preprocessing is an essential step in the machine learning workflow.

Correct Answer: True

K-means clustering requires the number of clusters to be specified in

advance.

Correct Answer: True

What is a primary advantage of decision trees?

A) They require a large amount of data.

B) They are easy to interpret and visualize.

C) They cannot handle categorical data.

D) They are always the most accurate model.

Correct Answer: B

Which technique is commonly used in unsupervised learning for grouping

similar data points?

A) Regression

B) Clustering

C) Classification

D) Reinforcement

Correct Answer: B

What is the purpose of the kernel trick in Support Vector Machines

(SVM)?

A) To minimize execution time

B) To reduce dimensionality

C) To enable non-linear separation of classes

D) To simplify data preprocessing

Correct Answer: C

Which ensemble method uses multiple decision trees to improve

accuracy?

A) Linear Regression
B) K-means

C) Random Forests

D) SVM

Correct Answer: C

In gradient boosting, how are new models trained?

A) Independently of previous models

B) To correct errors made by previous models

C) Using only labeled data

D) In parallel with existing models

Correct Answer: B

True/False Questions

The performance of KNN can degrade in high-dimensional spaces.

Correct Answer: True

Hierarchical clustering requires the user to specify the number of

clusters beforehand.

Correct Answer: False

Linear regression assumes a non-linear relationship between variables.

Correct Answer: False

DBSCAN is effective at identifying clusters of arbitrary shape.

Correct Answer: True

Feature selection is unnecessary if the dataset is large enough.

Correct Answer: False

What is a common application of reinforcement learning?

A) Image classification

B) Fraud detection

C) Game playing
D) Customer segmentation

Correct Answer: C

Which metric is commonly used to evaluate the performance of a

classification model?

A) Mean Squared Error (MSE)

B) R-squared

C) Accuracy

D) Precision-Recall Curve

Correct Answer: C

What does PCA stand for in data preprocessing?

A) Principal Component Analysis

B) Predictive Component Algorithm

C) Primary Cluster Analysis

D) Pattern Classification Algorithm

Correct Answer: A

Which of the following is an example of a regression task?

A) Classifying emails as spam or not

B) Predicting house prices

C) Grouping customers by purchasing behavior

D) Identifying objects in images

Correct Answer: B

What does the term "overfitting" refer to in machine learning?

A) A model that is too simple

B) A model that performs well on training data but poorly on unseen data

C) A model that has high bias

D) A model that generalizes well

Correct Answer: B

True/False Questions

Unsupervised learning relies on labeled data to identify patterns.

Correct Answer: False

Support Vector Machines are primarily used for regression tasks.

Correct Answer: False

Random Forests can help mitigate overfitting compared to single decision

trees.

Correct Answer: True

Data normalization is a step in data preprocessing that scales features to

a common range.

Correct Answer: True

K-means clustering can effectively handle non-spherical clusters.

Correct Answer: False

Which of the following methods is used for dimensionality reduction?

A) K-means clustering

B) Decision Trees

C) t-Distributed Stochastic Neighbor Embedding (t-SNE)

D) Support Vector Machines

Correct Answer: C

What is the primary purpose of feature engineering?

A) To clean the dataset

B) To create new features that enhance model performance

C) To visualize data

D) To select the best model

Correct Answer: B
In which scenario would you use ensemble methods?

A) When a single model performs well

B) To combine predictions from multiple models for improved accuracy

C) To simplify the model

D) When only linear relationships exist

Correct Answer: B

Which clustering algorithm is sensitive to noise and outliers?

A) K-means

B) DBSCAN

C) Hierarchical Clustering

D) Both A and C

Correct Answer: D

What is a key disadvantage of using KNN?

A) It is easy to interpret

B) It is computationally expensive during prediction

C) It requires labeled data

D) It works well in high dimensions

Correct Answer: B

True/False Questions

Feature selection helps reduce the dimensionality of the dataset.

Correct Answer: True

Linear regression can model relationships that are not linear.

Correct Answer: False

Reinforcement learning involves learning through feedback from actions

taken.

Correct Answer: True

The performance of SVM is not affected by the choice of kernel.

Correct Answer: False

Clustering algorithms can be used for both exploratory data analysis and
as preprocessing steps.

Correct Answer: True

What is the main purpose of feature scaling in machine learning?

A) To increase the amount of data

B) To standardize features to a comparable scale

C) To reduce model complexity

D) To enhance the interpretability of algorithms

Correct Answer: B

Which scaling method subtracts the mean and divides by the standard
deviation?

A) Min-Max Scaling

B) Normalization

C) Standardization

D) Robust Scaling

Correct Answer: C

What technique is used to measure the importance of features based on

their contribution to reducing impurity in Random Forests?

A) Information Gain

B) Fisher's Score

C) Feature Importance

D) PCA

Correct Answer: C

Which feature selection method starts with an empty model and adds
features one by one?
A) Backward Selection

B) Forward Selection

C) Exhaustive Selection

D) Regularization

Correct Answer: B

What is a key advantage of using PCA?

A) It increases the number of features

B) It captures variance in correlated features

C) It requires labeled data

D) It simplifies model interpretation

Correct Answer: B

True/False Questions

Feature selection can improve the predictive accuracy of a classification

algorithm.

Correct Answer: True

Wrapper-based feature selection methods do not rely on a specific

machine learning model.

Correct Answer: False

SMOTE is used primarily for generating synthetic samples for the

majority class.

Correct Answer: False

Regularization techniques can help prevent overfitting in machine

learning models.

Correct Answer: True

Data augmentation is only applicable to image datasets.

Correct Answer: False

Which technique involves creating new samples by interpolating between

existing samples of the minority class?
A) Random Sampling

B) SMOTE

C) PCA

D) k-Nearest Neighbors

Correct Answer: B

What does robust scaling use to normalize features?

A) Mean and standard deviation

B) Maximum and minimum values

C) Median and inter-quartile range

D) Mode and variance

Correct Answer: C

Which feature selection method evaluates all possible combinations of

features?

A) Forward Selection

B) Backward Selection

C) Exhaustive Feature Selection

D) Filter-Based Selection

Correct Answer: C

What is the primary goal of feature selection?

A) To increase the number of features

B) To improve model interpretability and performance

C) To generate new data points

D) To reduce computational cost without analysis

Correct Answer: B

Which of the following is NOT a method of feature scaling?

A) Min-Max Scaling
B) Standardization

C) Forward Selection

D) Robust Scaling

Correct Answer: C

True/False Questions

Information Gain helps measure how much knowing a feature decreases

uncertainty about the target variable.

Correct Answer: True

Random Noise Injection can be applied effectively to categorical features.

Correct Answer: False

L1 regularization can lead to some feature coefficients being reduced to

zero.

Correct Answer: True

ICA is primarily used for identifying correlated features.

Correct Answer: False

Cross-validation helps ensure a model generalizes well across different

subsets of the dataset.

Correct Answer: True

What is the main benefit of using Min-Max Scaling?

A) It centers data around the mean.

B) It transforms data into a fixed range, typically [0, 1].

C) It reduces the number of features.

D) It adds randomness to data.

Correct Answer: B

Which of the following techniques is used to measure the separation

between classes?

A) Information Gain
B) Fisher's Score

C) Random Forest Importance

D) PCA

Correct Answer: B

What is the purpose of adding a penalty term in regularization?

A) To increase the model complexity

B) To prevent overfitting

C) To reduce the size of the dataset

D) To enhance data augmentation

Correct Answer: B

Which method is ideal for handling datasets with many irrelevant

features?

A) L2 Regularization

B) L1 Regularization (Lasso)

C) Backward Selection

D) Exhaustive Selection

Correct Answer: B

What does PCA primarily aim to achieve?

A) Maximize the number of features

B) Capture maximum variance with fewer components

C) Separate independent components

D) Normalize data to a common scale

Correct Answer: B

True/False Questions

Feature scaling is unnecessary when all features are already on a similar

scale.
Correct Answer: True

Backward Selection starts with no features and adds them based on

performance improvement.

Correct Answer: False

Data augmentation can help improve a model's robustness to variations

in data.

Correct Answer: True

Regularization is only applicable to linear models.

Correct Answer: False

Independent Component Analysis (ICA) is effective for separating mixed

signals.

Correct Answer: True

What is the main goal of feature selection?

A) To minimize the dataset size

B) To improve model performance and interpretability

C) To increase the number of features

D) To create synthetic data

Correct Answer: B

Which scaling technique is specifically designed to be robust against

outliers?

A) Min-Max Scaling

B) Standardization

C) Robust Scaling

D) Normalization

Correct Answer: C

What is the primary function of k-Nearest Neighbors (k-NN) in data

augmentation?

A) To classify data points

B) To create synthetic samples based on existing neighbors

C) To reduce dimensionality

D) To normalize data

Correct Answer: B

Which feature selection technique evaluates features based on their

impact on a specific machine learning model?

A) Filter-Based Selection

B) Wrapper-Based Selection

C) Embedded Selection

D) Both A and B

Correct Answer: B

What is a significant drawback of exhaustive feature selection?

A) It is computationally efficient.

B) It can only be used with small datasets.

C) It requires labeled data.

D) It is prone to overfitting.

Correct Answer: B

True/False Questions

Synthetic Data Generation (SMOTE) is used to balance classes in

imbalanced datasets.

Correct Answer: True

Normalization ensures that the data has a mean of zero and a standard
deviation of one.

Correct Answer: False

Forward Selection evaluates features based on their individual

contribution to model performance.

Correct Answer: True

 Random Forest Importance can be used to assess the contribution of
individual features in a model.

Correct Answer: True

Principal Component Analysis (PCA) is a non-linear dimensionality

reduction technique.

Correct Ansiwer: False

Sabri Alhammadi i