Jimma University, Jimma Institute Of Technology

Faculty of Computing and Informatics(FCI)

Department Of Computer Science

COSC3112 – Introduction To Artificial Intelligence

Questionnaires for Exit Exam Preparation, June 2023, Prepared by Bekan Kitaw(M. Sc)
Instruction : Choose the correct answer.
1. Which of the following statements about the goals of Artificial Intelligence is incorrect:
A. Creating a system that mimics human cognitive abilities.
B. Establishing an intelligent connection between perception and action.
C. Replace humans and make them obsolete.
D. Building machines that can perform tasks that typically require human intelligence.
2. What is the main goal of the Act rationally approach to AI?
A. To develop machines that can mimic human behavior.
B. To develop machines that can learn from experience.
C. To develop machines that can act like humans.
D. To develop machines that can behave optimally in any given situation.
3. ABC Corporation is developing an AI system to predict stock prices. However, the stock market
is highly volatile and subject to frequent changes. What type of environment is the stock market
an example of?
A) Stochastic B) Non-observable C) Continuous D) Strategic
4. After implementing uniform cost search on the search tree shown on the right, and considering a
cost of 2 for each action, what would be the sequence of traversal in order to visit all the available
nodes starting from node A.
A) ABEFCDGH
B) ABCDEFGH
C) ABECDFGH
D) ADGHCBFE

5. On the above question #1 what would be the traversal order if we

perform greedy search with the heuristic function h(n) = 1 for all
nodes?

A) ABEFCDGH B) ABCDEFGH C) ABECDFGH D) ADGHCBFE

1
6. Imagine you are designing a robot that is tasked with exploring an unknown environment, and its
objective is to explore as much of the environment as possible. What search strategy would you
recommend for the robot?

A) Informed search such as IDS C) Informed search such as A* search

B) Uninformed search such as greedy search D) Uninformed search such as DFS

7. Suppose John is planning to create an AI agent that can assist a student in deciding which course
to take in the upcoming semester. This AI agent will make decisions based on a range of actions
that can be executed and a variety of possible outcomes for each action, each of which has an
assigned value. The agent will choose the action that maximizes the expected value, taking into
account its goals and knowledge of the current state of the world. Given this context, what type of
agent is John planning to build?
A) Model based agent C) Utility based agent
B) Simple reflex Agent D) Goal based agent

8. Imagine John is designing a machine learning algorithm that can recognize different types of fruit
in real-time. Which of the following is not among the steps that John would take to design and
implement this algorithm?

A. Data collection B. Feature extraction C. Model selection D. None.

9. Which of the following is an advantage of using informed search strategies in search problems?
A) They guarantee finding the optimal solution
B) They are always faster than uninformed search strategies
C) They can use domain-specific knowledge to guide the search
D) They are less complex than uninformed search strategies
10. Which of the following statements is true about the use of heuristics in search problems?
A) Heuristics are only used in artificial intelligence and computer science.
B) Heuristics are used to guide the search for a solution based on domain-specific knowledge.
C) Heuristics always guarantee finding the optimal solution to a problem.
D) Heuristics are only used in uninformed search algorithms like Breadth-First Search (BFS)
and Depth-First Search (DFS).
11. Which of the following best describes heuristic function used in heuristic search algorithms?
A) It is the number of nodes explored in the search tree.
B) It is the cost of the path from the start state to the current state.
C) It is the number of steps taken to reach the goal state.

2
 D) It is the distance or cost to the goal state from the current state.
12. Which of the following is a characteristic of heuristic search algorithms?
A) They explore the search space blindly.
B) They do not use any domain-specific knowledge to guide the search.
C) They always guarantee finding the optimal solution to a problem.
D) None of the above.
13. Which of the following search type is known by the use of heuristic search algorithm?
A) Breadth-First Search (BFS) C) A* search
B) Depth-First Search (DFS) D) Binary search
14. Which of the following is an example of an uninformed search algorithm?
A) A* search C) Map coloring
B) Iterative Deepening search D) Greedy search
15. In which scenarios is Iterative Deepening Search (IDS) preferred over other search algorithms?
A) When the search space is small and well-structured.
B) When the optimal solution needs to be found and memory usage is not a concern.
C) When there are constraints on the available memory for search.
D) When the search problem involves multiple goals or objectives.
16. Which of the following best explains the working principle of Iterative Deepening Search (IDS)?
A) IDS explores the search space breadth-first by using a queue to store the nodes.
B) IDS explores the search space depth-first and keeps track of the depth limit.
C) IDS uses heuristics to estimate the distance from the current node to the goal.
D) IDS uses a-priori knowledge to guide its search and avoid exploring irrelevant paths.
17. Which of the following is true regarding the domain-specific constraint used in the graph coloring
problem?
A) Each node must be assigned a color.
B) Each edge must be assigned a weight.
C) Each node must be connected to at least one other node.
D) Each node must have a certain degree.
18. Which of the following is not a common knowledge representation technique in AI?
A. First-order logic B. Semantic networks C. Neural networks D. Frames
19. Which of the following is an example of a first-order logic statement?
A) Three is an odd number. C) Water is composed of H and O2.
B) All birds can fly. D) All.
20. Which of the following is a valid sentence?

3
 A. P ∧ Q → P C. ( ( ( P ∨ Q ) ∧ ¬ Q ) → P )
B. P → Q ↔ Q → P D. All except B.
21. Which of the following statements is satisfiable?
A) P∧Q∧R∧¬P B) P∨Q∨R∨¬P C) P→Q→R→¬P D) P↔Q↔R↔¬P
22. What is an inference engine in a rule-based system?
A) The component that stores the rules
B) The component that provides an interface for humans to interact with the system
C) The component that applies the rules to a set of facts
D) The component that visualizes the system's reasoning
23. In a frame-based system, what is an attribute?
A) A characteristic of a frame that defines its relationship to other frames
B) A property of a frame that describes its properties or characteristics
C) The frame's location in a hierarchical structure
D) The rule that relates a frame to other frames
24. Which of the following is a knowledge representation technique used in Artificial Intelligence that
represents knowledge in the form of objects and their properties?
A) Semantic Networks B) Frames C) Production Systems D) Ontologies
25. What type of learning involves learning from interacting with an environment and receiving
feedback in the form of rewards or penalties?
a. Supervised Learning c. Reinforcement Learning
b. Unsupervised Learning d. None of the above
26. Which of the following is an example where reinforcement learning is being employed?
a) Training a chatbot to understand and respond to user queries.
b) Analyzing customer data to identify market segments.
c) Teaching a robot to navigate a maze and find the shortest path.
d) Clustering news articles based on their content.
27. What is the main difference between Reinforcement Learning and Unsupervised Learning?
A) Reinforcement Learning uses rewards and punishments, while Unsupervised Learning uses
labeled data.
B) Reinforcement Learning uses clustering algorithms, while Unsupervised Learning uses
rewards and punishments.
C) Reinforcement Learning learns from interactions with an environment, while Unsupervised
Learning finds patterns in unlabeled data.

4
 D) Reinforcement Learning uses labeled data, while Unsupervised Learning uses rewards and
punishments.
28. Which type of supervised learning is used when the target variable is categorical?
a) Classification b) Regression c) Clustering d) Reinforcement Learning
29. Which task is associated with classification in supervised learning?
a) Predicting a continuous value
b) Grouping data into predefined categories
c) Identifying patterns in unlabeled data
d) None of the above
30. Which algorithm is commonly used for classification tasks?
a) Linear regression c) K-means clustering
b) Decision tree d) Association mining
31. Which algorithm is commonly used for regression tasks?
a) Logistic regression c) Support vector Machines
b) K-means clustering d) Bayesian Linear Regression
32. What is the primary objective of regression in supervised learning?
a) To predict a continuous value c) To find patterns in unlabeled data
b) To group data into predefined categories d) To optimize a cost function
33. Which learning approach is commonly used when the target variable is numerical?
a) Classification c) Clustering
b) Regression d) Reinforcement Learning
34. Association mining is a technique used to:
a) Predict future values of a continuous variable
b) Identify patterns and relationships in large datasets
c) Classify data into predefined categories
d) Cluster similar data points together
35. The goal of association mining is to discover:
a) Cause-and-effect relationships
b) Correlations and associations between items
c) Optimal clustering solutions
d) Anomalies in the data
36. Association mining is often applied to analyze data in which domain?
a) Healthcare b) Finance c) Retail d) All of the above
37. Which of the following is an example of Association Mining?

5
 a. Predicting the price of a stock
b. Classifying emails as spam or not spam
c. Recommending products to customers based on their purchase history
d. Identifying the sentiment of customer reviews
38. In linear regression, the relationship between the dependent variable and independent variables is
assumed to be:
a) Linear b) Non-linear c) Exponential d) Logarithmic
39. The output of linear regression is:
a) Probabilities c) Continuous values
b) Binary outcomes d) Categorical labels
40. Which of the following supervised learning technique creates decision boundary to segregate n-
dimensional space into classes?
a) Linear Regression c) Decision Trees
b) Logistic Regression d) Support Vector Machine
41. Linear regression is a supervised learning algorithm used for:
a) Classification c) Regression
b) Clustering d) Dimensionality reduction
42. Technique used by social network such as Netflix to recommend the movies and web-series to its
users as per the watch history using unlabeled data would be categorized in which ML algorithm?.
A. Clustering B. Regression C. Classification D. Ensembling
43. Which of the following neural networks is commonly used for image recognition tasks?
a) Convolutional Neural Network (CNN) c) Feedforward Neural Network (FNN)
b) Recurrent Neural Network (RNN) d) Deep Belief Network (DBN)
44. All of the following options are limitations of CNNs in AI EXCEPT:
A. Decisions are based on the current input.
B. They work based on saving the output of a layer and feeding it back as input to predict the
output of the layer.
C. They do not have memory.
D. A and C.
45. A social media company wants to predict whether a user will click on an advertisement. They have
a dataset of user demographics and past advertisement click history. They decide to use a Support
Vector Machine (SVM) algorithm to make the prediction. What is the goal of the social media
company in the above case study?
a) To predict user demographics

6
 b) To predict user advertisement click history
c) To predict whether a user will click on an advertisement
d) To group similar users together
47. A marketing agency wants to determine the effectiveness of different marketing campaigns in
generating leads. They collect data on the type of campaign (email, social media, search ads), the
number of leads generated, and other relevant factors. They decide to use linear regression to
analyze the relationship between campaign type and lead generation. What is the goal of this case
study?
a) Predicting customer churn
b) Forecasting stock prices
c) Analyzing the effectiveness of marketing campaigns in generating leads
d) Estimating population parameters
48. A healthcare company wants to develop a chatbot that can answer patient queries and provide
medical advice. Which application of NLP would be most suitable for this task?
a) Sentiment analysis c) Question answering
b) Named entity recognition d) Machine translation
49. A news organization wants to automatically categorize news articles into different topics such as
sports, politics, and entertainment. Which application of NLP would be most suitable for this task?
a) Text summarization c) Text Recognition
b) Topic modeling d) Named entity recognition
50. A company wants to develop a virtual assistant that can understand and respond to voice
commands. Which application of NLP would be most suitable for this task?
a) Speech recognition c) Text summarization
b) Sentiment analysis d) Named entity recognition
51. A researcher is working on a project to analyze social media data. They want to reduce words to
their base or root form to simplify the analysis and capture the essence of the text. Which NLP
technique would be most suitable for this task?
a) Tokenization b) Stemming c) Lemmatization d) None.
52. A language processing system is being developed to improve search engine results. The system
needs to convert words into their canonical form to ensure accurate matching and retrieval of
relevant documents. Which NLP technique would be most suitable for this task?
a) Tokenization b) Stemming c) Lemmatization d) None.

7
53. A company is developing a chatbot for customer support. They want to preprocess user queries by
breaking them down into individual words to analyze the intent and provide appropriate responses.
Which NLP technique would be most suitable for this task?
a) Tokenization b) Stemming c) Lemmatization d) None.
54. A company wants to analyze customer feedback to identify the most frequently mentioned
products and their associated sentiments. Which NLP technique would be most suitable for
extracting product names?
a) POS tagging c) Stop word removal
b) Named Entity Recognition d) Tokenization
55. A company wants to analyze customer reviews to identify the parts of speech used in positive and
negative sentiments. Which NLP technique would be most suitable for this task?
a) POS tagging c) Stop word removal
b) Named Entity Recognition d) Tokenization
56. A marketing agency wants to determine the effectiveness of different marketing campaigns in
generating leads. They collect data on the type of campaign (email, social media, search ads), the
number of leads generated, and other relevant factors. They decide to use linear regression to
analyze the relationship between campaign type and lead generation. What is the goal of this case
study?
a) Predicting customer churn
b) Forecasting stock prices
c) Analyzing the effectiveness of marketing campaigns in generating leads
d) Estimating population parameters
57. An AI-powered recruitment system consistently rejects applications from female candidates, even
when their qualifications are comparable to male candidates. This is an example of:
a) Bias in AI b) Discrimination in AI c) A and B d) None of the above
58. An AI chatbot consistently responds to derogatory or offensive user inputs with inappropriate or
offensive language. This is an example of:
a) Bias in AI
b) Discrimination in AI
c) Both bias and discrimination in AI
d) Inadequate training data

8
59. An AI-powered criminal justice system disproportionately recommends longer sentences for
individuals from marginalized communities compared to others with similar criminal histories.
This is an example of:
a) Bias in AI c) Both bias and discrimination in AI
b) Discrimination in AI d) None of the above
60. An AI-powered loan approval system denies loans to individuals from certain ethnic backgrounds,
even when they meet all the necessary criteria. This is an example of:
a) Bias in AI
b) Discrimination in AI
c) Both bias and discrimination in AI
d) None of the above

Note that:

· Please be aware that the questions provided cover all the chapters from Chapter One to

Chapter Eight. I hope that these questions are beneficial, and in case you encounter any

difficulties, please don't hesitate to reach out to me for assistance.

· It's important to note that these questions are specifically tailored for Computer Science

students. However, there may be some variations for students in other departments such

as Software Engineering, Information Technology, and Information Sciences.