You are on page 1of 2

COMP 456: Artificial Intelligence and Expert Systems

Sample Exam

Question 1:
Given two propositions P and Q, the logical operator “↔“ is read equivalent to. P ↔ Q is
defined as being equivalent to (P → Q) ˄ (Q → P).
Based on this definition show that P ↔ Q is logically equivalent to (P → Q) ˄ (Q → P).
Use a truth table to explain your answer.

Question 2:
Contrast the use of data-driven strategy versus goal driven strategy for search control in
production systems.

Question 3:
Write a program that reads a list of integers Y, and an integer X and finds if X = the
multiplication of elements of Y.

Question 4:
Explain using an example how the best-first heuristic principle is used in the A*
algorithm.

Question 5:
Define inductive reasoning. Give an example to explain it.

Question 6:
Consider the following set of rules:

1: IF it is raining AND the temperature is cold


THEN playground is empty

2: IF kids are playing videogame


THEN kids are not playing outside

3: IF kids are not playing outside


THEN kids are playing videogames or kids are at school

4: IF it is not a school day


THEN kids are playing outside OR kids are playing videogames

5: IF kids are not playing outside


THEN playground is empty

Use rules 1,2,3,4, and 5 to draw a tree to show how an inference engine uses depth-first
search strategy to prove that ‘playground is empty’ given that “it is not a school day” and
“kids are playing videogame”. (Use the rules as the nodes of the tree and give them
numbers to show their sequence of trial.)

Question 7:
List three advantages of each of the following reasoning approaches for implementing
expert systems:
- Rule-based reasoning,
- Case-based reasoning,
- Model-based reasoning.

Question 8:
Consider the following statements:
- When the cat is absent, the mousses are present.
- When the dog is present, the cat is absent.
- When the cat is not sleeping, he is absent or present.
- When the dog is absent and the cat is absent, the mousses are presents.
- The dog is absent.
- The cat is not sleeping.
a- Translate these statements into a knowledge base form.
b- Using a forward chaining strategy, list all possible facts that can be derived from this
knowledge base. Note that every new derived fact is added to the knowledge base and
will be used for deriving other new facts. The reasoning process is recursively used
until no new fact is derived.

Question 9:
Briefly explain how the back-propagation algorithm works for the multilayer neural
network.

You might also like