W1-2-60-1-6

JOMO KENYATTA UNIVERSITY OF AGRICULTURE AND TECHNOLOGY

UNIVERSITY EXAMINATIONS 2023/2024
BACHELOR OF SCIENCE IN INFORMATION TECHNOLOGY/BACHELOR OF SCIENCE
IN BUSINESS COMPUTING
BIT 2319 ARTIFICIAL INTELLIGENCE
SEPTEMBER-DECEMBER 2023 ORDINARY EXAMINATIONS
DATE: DECEMBER 2023 TIME: 2 HOURS

INSTRUCTIONS:
Attempt Question 1 and any other two questions.
1)
a) Define Artificial in terms of rational thinking [2 Marks]
b) Discuss what Al can do today,citng appropriate examples [4 Marks]c)
i) Give three examples of real world problems [3 Marks]
d) [4 Marks]ii) Explain the components well-defined
problems in Artificial Intelligence
i) List the four basic types of agent program in any intelligent system [4 Marks]
c) [3 Marks]ii) Explain how you can convert them into
learning agents
i) Differentiate an agent function and an agent program [2 Marks]
ii) Explain why a goal-stack plannur would have difficulty finding a good solution to
the following blocks world problem? (Assume the "standard" operators of pickup,
putdown,stack,and unstack) [4 Marks]

on(B,C), on(D, A)
Initial state Goal State
f) Differentiate between list, tuple, set and dictionary in Python programming with one
example each [4 Marks]
2)
a) Differentiate uninformed search (blind search) and informed search (heuristic
Search)strategies [4 Marks]
b) Explain the criteria to measure the performance of different search strategies [4 Marks]
c) Define the following search strategies:
i) Best-first-search. [2 Marks]
ii) A*Search [2 Marks]
d) Consider the search graph below, where S is the start node and G1, G2, and G3 are goal
states.Arcs are labeled with the cost of traversing them and the heuristic cost to a goal is
shown inside the nodes. For each of the three search strategies below, indicate which of the
goal states is reached using appropriate diagrams, and show order of node expansion:
 (A) (B) (C)
d) Consider the hypothetical graph below of Predictive Error (y-axis) vs. Model
Complexity(x-axis), and how Error on Test/Training Data varies as Model Complexity
increases.

Predictive
Error

Model Complexity

Label the following regions as A, B, or C, depending on whether they correspond

to Overfitting, Underfitting, or Ideal Model Complexity:
i) Overfitting (Write A,B,or C) [1 Mark]
ii) Under fitting (Write A,B,or C) [1 Mark]
iii) Ideal Model Complexity (Write A, B, or C) [1 Mark]

i) Breadth-first search.
ii) Uniform cost search.
iii)A*search.
3) [8 Marks]
 a) Explain using appropriate examples expert system shells and tools [4 Marks]b)
i) Outline the elements of propositional logic ii) Use truth tables to show the validity of the
following distributive law: [2 Marks]PA(QVR)→(PAQ)V(PAR)c) Consider the following
sentences: [4 Marks]
i) John likes all kinds of food
ii) Apples are food
iii) Chicken is food
iv) Anything anyone eats and isn't killed by is food
v) Bill eats peanuts and is still alive
vi) Sue eats everything bill eats
(1) Translate these sentences into formulas in predicate logic [6 Marks]4)
[4 Marks](2) Prove that john likes peanuts using backward chaining

i) Discuss the different types of sources of data from where we can collect reliable and
authentic datasets ii) Explain the data exploration stage [4 Marks]
b) Explain the following Machine Learning Learning algorithms: [3 Marks]i) K-
Nearest Neighbors
ii) Linear Regression
iii) XGBoost
iv) Support Vector Machines
[4
Marks]c) Match each decision boundary below with the classifier type most likely to produce
such a
i) Decision Tree Classifier.(Write A,B,or C) [2 Marks]
iii)
Near
est
Neig
hbou
r
Classi
fier.
(Writ
e
A,B,o
r C)
[2
Mark
s][2
Mark
s]