ESCAE UNIVERSITY, BENIN

Course:Artificial Intelligence Exam Code: COSC 415 Time Allowed : 90min

Instruction: Answer all questions

_NAME : ______________________________________MATRIC NO:__________

1. What is the main goal of Artificial 9. Which type of AI system can

Intelligence (AI)? a) To solve problems operate independently of human
in mathematics b) To simulate human intervention? a) Weak AI b) Narrow AI
intelligence c) To create artistic works d) c) Strong AI d) None of the above
To build faster computers
10. Which technique is used to reduce
2. Which of the following is a subset of the dimensions of a dataset? a)
AI that focuses on data analysis and Clustering b) Principal Component
pattern recognition? a) Machine Analysis (PCA) c) Regression d)
Learning b) Robotics c) Natural Classification
Language Processing d) Expert Systems

3. Who is considered the father of AI? 11. What is overfitting in the context
a) Alan Turing b) John McCarthy c) of machine learning? a) When a model
Marvin Minsky d) Geoffrey Hinton performs well on unseen data but poorly
on training data b) When a model
4. Which of the following AI performs well on training data but
techniques is used for speech poorly on unseen data c) When a model
recognition? a) Computer Vision b) performs equally well on both training
Natural Language Processing c) and unseen data d) When a model
Robotics d) Expert Systems underestimates the complexity of the
data
5. What is an artificial neural network
(ANN) inspired by? a) Quantum 12. Which programming language is
mechanics b) Human brain c) Classical most commonly used for AI
physics d) Relativity theory development? a) Java b) C++ c) Python
d) Ruby
6. Which type of learning involves an
agent interacting with an environment
to maximize some notion of 13, What is the main function of a loss
cumulative reward? a) Supervised function in machine learning? a) To
Learning b) Unsupervised Learning c) compute the accuracy of the model b)
Reinforcement Learning d) Semi- To measure the difference between
supervised Learning predicted and actual values c) To
optimize the model parameters d) To
7. In AI, what does NLP stand for? a) split the dataset into training and testing
Neuro-Linguistic Programming b) sets
Natural Language Processing c) Neural
Learning Protocol d) None of the above
14. Which type of neural network is
8. Which algorithm is commonly used typically used for image recognition?
for classification and regression tasks? a) Recurrent Neural Network (RNN) b)
a) K-means b) Decision Tree c) Genetic Convolutional Neural Network (CNN) c)
Algorithm d) Ant Colony Optimization Generative Adversarial Network (GAN)
d) Feedforward Neural Network (FNN) Decision Tree b) Random Forest c)
Generative Adversarial Network (GAN)
15. What is the purpose of a Turing d) Support Vector Machine
Test? a) To evaluate the speed of a
computer b) To assess the creativity of 21 Which technique is used to handle
an AI system c) To determine if a missing data in a dataset? a)
machine can exhibit intelligent behavior Normalization b) Imputation c)
equivalent to a human d) To measure the Standardization d) Binarization
accuracy of AI algorithms

16. Which of the following is an 22/ What does the term 'bias' refer to
example of unsupervised learning? a) in the context of machine learning? a)
Linear Regression b) Support Vector The difference between the expected and
Machines c) K-means Clustering d) actual output b) The tendency of a
Random Forest model to favor certain outcomes c) The
complexity of a model d) The ability of
17. What is a chatbot? a) A type of a model to generalize to new data
robot that can walk and talk b) A
program designed to simulate
conversation with human users c) A 23. Which of the following is a method
system for managing large datasets d) A for feature selection? a) Principal
method for optimizing machine learning Component Analysis b) Gradient
models Descent c) Hyperparameter Tuning d)
Cross-validation
18. Which of the following is NOT a
type of machine learning? a) 24. In reinforcement learning, what is
Supervised Learning b) Unsupervised the 'reward signal'? a) The action
Learning c) Reinforcement Learning d) taken by the agent b) The feedback
Heuristic Learning received from the environment c) The
state of the environment d) The policy
19. What is the main advantage of followed by the agent
using deep learning over traditional
machine learning? a) Requires less data
b) Can automatically learn features from 25/ Which of the following is NOT an
raw data c) Is less computationally application of AI? a) Autonomous
intensive d) Provides deterministic Vehicles b) Weather Forecasting c)
results Personal Assistants (e.g., Siri, Alexa) d)
Classical
20. Which of the following is an
example of a generative model? a)
Mechanics

Instruction: Answer only two questions in this section

Question 1
Define Artificial Intelligence. List five tasks that computers are unlikely to be able to do
within the next five years

Question 2
Describe the salient features of an agent.What is an autonomous agent

Question 3
Compare and contrast supervised learning and unsupervised learning in the context of
artificial Intelligence. Provide examples of real-world appications where each type of learning
is most suitable, and discuss the challenges associated with each approach