Cairo University

Faculty of Computers and Information

Final Exam

Department: Computer Sciences

Course Title: Artificial Intelligence

Course Code: CS361
Semester: Spring 2015/2016
Instructors: Dr. Khaled Wassif
Dr. Hanaa Bayomi
Date: 25 May 2016
Exam Duration: 2 Hours

Question Mark Signature

One
Two
Three
Four
Five
Six
Seven

60 Eight
Nine

Total Marks

Total Marks in Writing:

Attempt ALL Questions:
Part (1)
Q1) Mark each statement with T or F in the right side: (10 Marks)
1. Rational agents maximize actual performance, while perfect agents ( )
maximize expected performance. ( F )
2. Table-driven agents are adequate for solving most practical ( )
problems. ( F )
3. Model-based reflex agents select actions on the basis of the current ( )
percept with ignoring the rest of the percept history. ( F )
4. Simple reflex agents work only if the environment is fully observable. ( )( T )
5. In some cases, a goal-based agent is inefficient but a utility-based ( )
agent can still make rational decisions.
6. With problem-solving agents (except contingency problems), a ( )
solution is executed without paying attention to percepts.
7. The memory requirements of a search strategy are a bigger problem ( )
than its execution time. ( T )
8. An informed search strategy can find solutions more efficiently than ( )
an uninformed strategy. ( T )
9. Depth-first search strategy is complete but not optimal. ( )( F )
10. Best-first search is implemented by ordering the nodes in fringe in ( )
descending order of evaluation. ( F )
11. Greedy search takes the biggest bite possible to reach the goal ( )
without worrying about whether this will be best in the long run. ( T )
12. An admissible heuristic function is overestimating the cost to reach ( )
the goal. ( T )
13. A problem with more restrictions on the actions is called a relaxed ( )
problem. ( F )
14. In many optimization problems, the path to the goal is irrelevant; the ( )
goal state itself is the solution.
 15. Simulated annealing search frequently gets stuck when reaches a ( )
point at which no progress is being made.
16. A local maximum is a peak that is lower than each of its neighboring ( )
states but higher than the global maximum.
17. A knowledge-based agent can be described at two levels: knowledge ( )
level and logical level.
18. Logical entailment is based only on syntax and does not depend at all ( )
on logical inference. ( F )
19. An inference rule is complete if the conclusion is true in all cases ( )
where the premises are true. ( F )
20. Resolution is complete for the propositional logic. ( )
Q2) The first step to design an agent is to specify its task environment
as fully as possible. Explain briefly the main dimensions that
categorize the task environments. Accordingly, what are the
characteristics of Wumpus world environment? (9 Marks)
Q3) Describe the basic components that formally define the problem of
a problem-solving agent and show the relation between these
components and its state space. (8 Marks)

Q4) Describe and evaluate A* search strategy. What are the difference
between A* and IDA*? (8 Marks)
Q5) Decide whether each of the following sentences is valid, satisfiable,
or unsatisfiable. Verify your decisions using model checking.
(5 Marks)
a. Smoke ⇒ Fire

b. Smoke  Fire  Fire

c. (Smoke ⇒ Fire) ⇒ (Smoke ⇒ Fire)

d. (Smoke ⇒ Fire) ⇒ ((Smoke  Heat) ⇒ Fire)

e. ((Smoke  Heat) ⇒ Fire)  ((Smoke ⇒ Fire)  (Heat ⇒ Fire))

 Part (2)
Q6) in the game tree below, the numbers indicate values for the
heuristic evaluation function (in depth search 3 (figure 1) and in
depth search 4 (figure 2)). (9 marks)

Figure 1 Figure 2

a) Should Max choose the left or right branch if the depth of the search is 3 (i.e.,
Max chooses, Min chooses, Max chooses)? Show the minimax values in the tree.
(2 marks)

b) Should Max choose the left or right branch if the depth of the search is 4? Show
the minimax values in the tree. (2 marks)

c) Is the minimax values change when computed via a depth-3 vs. depth-4 search ,
Explain why. (2 marks)
 d) Using the full tree (figure 2), show alpha-beta pruning by circling all pruned
nodes. Assume that nodes are evaluated from left to right. (3 marks)

Q7) (7 marks)
a) Translate into first-order logic the following sentences:
(Remember to define all predicates, functions, and constants.)
(2 marks)
1- If someone understands someone, then he/she is that someone’s friend.

2- All brown horses are slow.

b) Translate the following logical statement into good, natural
English (no x’s and y’s): (2 marks)
∀ x, y, l SpeaksLanguage(x, l) ∧ SpeaksLanguage(y,l) → Understands(x, y) ∧
Understands(y, x)

c) For each pair of literals below, specify a most general unifier, or

indicate that they are not unifiable. (3 marks)
1) r(f(x), y) and r(z, g(w))

2) r(f(x), x) and r(y, g(y))

3) r(a, john, a) and r(f(x), x, y)

Q8) The graph below is a constraint graph for a CSP that has only
binary constraints. Initially, no variables have been assigned. For
each of the following scenarios, mark all variables for which the
specified filtering might result in their domain being changed.
(4 marks)

1)A value is assigned to A. Which domains might be changed as a result of running

forward checking for A?
( )A ( )B ( )C ( )D ( )E ( )F ( ) No one

2)A value is assigned to A, and then forward checking is run for A. Then a value is
assigned to B. Which domains might be changed as a result of running forward
checking for B?
( )A ( )B ( )C ( )D ( )E ( )F ( ) No one

3)A value is assigned to A, and then forward checking is run for A. Then a value is
assigned to D. Which domains might be changed as a result of running forward
checking for D?
( )A ( )B ( )C ( )D ( )E ( )F ( ) No one

4)A value is assigned to A, and then forward checking is run for A. Then a value is
assigned to C. Which domains might be changed as a result of running forward
checking for C?
( )A ( )B ( )C ( )D ( )E ( )F ( ) No one

With Our Best Wishes