You are on page 1of 7

FACULTY OF INFORMATION TECHNOLOGY

BACHELOR IN BUSINESS INFORMATION TECHNOLOGY


END OF SEMESTER EXAMINATION
BBT 3201: INTRODUCTION TO ARTIFICIAL INTELLIGENCE CONCEPTS
DATE: 21st August 2019 TIME: 2 Hours
Instructions:
1. This examination consists of FIVE questions.
2. Answer Question ONE (COMPULSORY) and any other TWO questions.
3. For multiple choice questions,marks are awarded ONLY for total correct answers.

Question ONE
a) Choose the options that are correct regarding machine learning (ML) and artificial
intelligence (AI), (2 marks)
i. ML is an alternate way of programming intelligent machines.
ii. ML and AI have very different goals.
iii. ML is a set of techniques that turns a dataset into a software.
iv. AI is a software that can emulate the human mind to an extent.
b) Which of the following sentence is FALSE regarding regression? (2 marks)
i. It relates inputs to outputs.
ii. It is used for prediction.
iii. It may be used for interpretation.
iv. It discovers causal relationships.
c) For the one-parameter model, mean-Square error (MSE) is defined as follows:
We have a half term in the front because, (2 marks)
i. Scaling MSE by half makes gradient descent converge faster.
ii. Presence of half makes it easy to do grid search.
iii. It does not matter whether half is there or not.
iv. None of the above
d) Distinguish between the following;
i. The law of thoughts and rational agent approach to AI. (2 marks)
ii. Deterministic and stochastic task environments. (2 marks)
iii. Feed-forward and recurrent neural networks. (2 marks)
____ Question 1 continues on page 2 ____

Page 1 of 4
e) Consider an agent that inhabits a finite planar maze world consisting of discrete cells.
Some pairs of neighboring cells have walls that prohibit motion between the cells. The
agent is located in exactly one cell at any point in time. The agent uses four actions to
navigate the world N, S, E, W , each moves to the adjacent cell in the associated compass
direction. The agent has two types of sensors:(i) A primitive 1-bit cheese sensor which
reads 1 if and only if the agent is in a block containing a piece of cheese; (ii) A wall
detector which each provide a reading of either Blocked or Open in each of the four
directions. There is a piece of cheddar in one (unknown) cell, but which can be reached
from the agent’s initial location.
i. Draw the structure of the agent described above. (3 marks)
ii. The agent above is considered ‘fragile’ in the sense that it cannot operate well in the
real world. Discuss two modifications to the structure that would overcome this
challenge. (4 marks)
f) Explain two computing paradigms that have resulted from the prominence of deep
learning in the recent past. (4 marks)
g) Agriculture is the largest contributor to the GDP of most east African countries. Discuss
two AI applications that can help to improve the productivity in this sector. (4 marks)
h) AI is about building models that simulate intelligence, these models are based on generic
problem solving approaches that we use in our daily lives. List one AI domain and discuss
one such approach applied in this domain. (3 marks)
(Total: 30 marks)

Question TWO
Consider the figure below representing choices in a two-player game. Perform the minimax
algorithm to establish the choice made by the current player. Please workout your solution on
the attached page which has a copy of the graph.
a) Without α-β pruning. (5 marks)
b) With α-β pruning. (10 marks)

(Total: 15 marks)

Page 2 of 4
Question THREE
After receiving yet another “Dear sir or madam..” email, you decide to construct a spam filter.
For your first attempt you decide to try using a k Nearest Neighbors model. You decide to
classify spam according to 2 features: email word count and occurrences of the words “sir” or
“madam”.

a) Draw the decision boundary for 1-nearest-neighbor on the above diagram(see attached
paper with diagram) of the given training data. Use the center of the faces as the positions
of the training data points. (13 marks)
b) How will 1-nearest-neighbor classify an email with 200 words of which 9 are the word
“sir”? (2 marks)
(Total: 15 marks)

Question FOUR
a) Consider the following Bayesian network where N = having pneumonia and F = fever.The
domain of N and F is {+,-}.Compute P(F | N=-) using inference by enumeration and
Bayes rule. (8 marks)

b) Consider the factorized joint probability


P(A,B,C,D) = P(D|B,C) x P(C|A) x P(B|A) x P(A)
____ Question 4 continues on page 4 ____

Page 3 of 4
i. Draw the corresponding Bayesian network. (2 marks)
ii. State whether the following are true given the bayes net in i above.
i. A⊥D (1 mark)
ii. B ⊥ C | A (2 marks)
iii. B ⊥ C | D (2 marks)
(Total: 15 marks)

Question FIVE
a) Consider the following state space graph with an initial state of A and goal G. The
heuristics are given in the table below.

state h(n)
A 6
B 3
C 4
D 3
E 3
F 4
G 0
H 1

i. Using a queue, show the path given by BFS. (7 marks)


ii. Solve (a) using the A* algorithm. Use a closed set and generate successors and break
ties in alphabetic order. (8 marks)
(Total: 15 marks)

_____ End of Examination _____

Page 4 of 4
Question TWO
Without α-β pruning.
Question TWO
With α-β pruning.
Question THREE

You might also like