Afdeling B. / Section B

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

Afdeling B. / Section B.

Beantwoord die volgende vrae in die antwoordboek en die antwoordblad wat verskaf is.
Answer the following questions in the answer script and on the answer sheet provided.

Vraag 1. / Question 1. (SU1) [15]

Gegewe die volgende definisies van Kunsmatige Intelligensie, beantwoord die vrae wat volg:
Given the following definitions of Artificial Intelligence, answer the questions that follow:
Definition 1: The goal of AI is to develop machines that behave as though they were intelligent.
- John McCarthy, 1955
Definition 2: Artificial Intelligence is the study of how to make computers do things at which, at the moment,
people are better. - Elaine Rich, 1983
Definition 3: AI is the ability of digital computers or computer-controlled robots to solve problems that are
normally associated with the higher intellectual processing capabilities of humans …
- Encyclopedia Britannica, 1991
1.1 Vir elkeen van die gegewe definisies, identifiseer die vaagheid en lewer kommentaar oor die probleme
wat geassosieer word met daardie vaagheid.
For each of the given definitions, identify the elements of vagueness and comment on the problems
associated with that vagueness. (9)
1.2 Wat is die verband tussen Kunsmatige Intelligensie en Breinwetenskap?
How does Artificial intelligence relate to Brain Science? (3)
1.3 Wat is die verband tussen Kunsmatige Intelligensie en Kognitiewe Wetenskap?
How does Artificial intelligence relate to Cognitive Science? (3)
1.1 Vagueness in Def. 1:
“intelligent” is a vague term because it is also difficult to define
- therefore it is difficult to have a test for intelligence
- e.g. random actions of robotic vehicles may appear intelligent but it is not
- complex behaviour can be produced by very simple electrical circuits
Vagueness in Def. 2:
- “at the moment” – implies that as computers get better at tasks, that is no longer AI.
- “People are better” – which people?
- “People are better” – how exactly do you measure “better” for complex tasks?
Vagueness in Def. 3:
Some “higher intellectual capability of humans” can easily be automated:
- computer with a large memory that can save a long text and retrieve it on demand
- computers can do difficult calculations faster than humans
- According to this definition, then, every computer is an AI system.
1.2 Through the research of intelligent systems we can try to understand how the human brain works
and then model or simulate it on the computer.
Many ideas and principles in the field of neural networks stem from brain science
with the related field of neuroscience.
1.3 Cognitive science is devoted to research into human thinking at a somewhat higher level.
Similarly to brain science, this field furnishes practical AI with many important ideas.
On the other hand, algorithms and implementations lead to further important conclusions
about how human reasoning functions.

CMPG313 5
Vraag 2. / Question 2. (SU2) [10]

2.1 Give 2 examples of the use of artificial intelligence in each of the following fields:
Logistics, Transportation, Medical. (6)
Note: Answers should indicate the “use of AI” to get marks.
For example, the following will not get any marks:
planes; electric cars; eliminate greenhouse gasses; delivery trucks; space ships; X-ray machines;
medical scanners; interactive machines; patient information systems; machines giving medication;
remote control machines; machines that detect viruses; war machines; MRI machines; computers in
laborotories or hospitals.

Answer examples:
Logistics:
(1) Navigation System (GPS) doing route planning with heuristic search.
(2) A courier company uses a software agent to plan an optimal delivery route for a single truck, taking
into account the shortest travel distance for travel cost or time as well as the priority of packages.
(3) An airline management system plans crew routes, crew rotation, risk, resource requirements
needed at airports for fuel, meals using information that changes in real-time such as weather,
cancellations, emergencies and other delays.

Transportation:
(1) self-driving cars / taxis / delivery trucks
(2) plane auto-pilot
(3) pilot training with profiling through machine learning in flight simulators

Medical:
(1) text retrieval and analysis of published medical research to help in drug discovery
(2) diagnosis of diseases using symptoms and test results
(3) pattern recognition in medical imaging

2.2 What are “expert systems” and what methods of artificial intelligence are used in them? (4)
Answer: Expert systems are knowledge-based systems where the “expertise” of human experts is encoded
as facts and rules into a knowledge base. Some form of logic-based inference mechanism can then use
the knowledge base to analyse problems and give advice to users. Uncertainty can be handled by fuzzy
logic or probabilities. Methods of decision tree learning from data also work with probabilities. Systems like
CART, ID3 and C4.5 can quickly and automatically build very accurate decision trees which can represent
propositional logic concepts and then be used as expert systems.

Vraag 3. / Question 3. (SU3) [15]

3.1 Volgens die sintaks van proposisionele logika, watter van die atribute (a tot d) hieronder is waar vir die
volgende formule? Verduidelik u antwoord. ((P → Q) ∧ Q) → P
According to the syntax of propositional logic, which of the attributes (a to d) below are true for the
following formula? Explain your answer. ((P → Q) ∧ Q) → P
(a) goed gevorm (a) well formed
(b) geldig (b) valid
(c) vervulbaar / bevredigbaar (c) satisfiable
(d) onvervulbaar (d) unsatisfiable (4)
Answer:
The sentence is (a) well-formed; (c) satisfiable.
A model for the sentence is (P=True, Q=False) so it is satisfiable.
The model (P=False, Q=True) makes the sentence false, so it is not valid.
Die sin is (a) goed gevorm; (c) bevredigend.
'n Model vir die sin is (P = Waar, Q = Onwaar) so dit is vervulbaar / bevredigbaar.
Die model (P = Onwaar, Q = Waar) maak die sin onwaar, dus is dit nie geldig nie.

CMPG313 6
3.2. Propositional logic resolution

3.2.1 Bou ‘n KB van proposisionele logika sinne vir die Engelse sinne hieronder.
Die finale vorm van elkeen moet CNF wees.

Construct a KB of propositional logic sentences for the English statements below.


The final form of each one should be CNF. (6)

Intermediate steps Answer in CNF:

Either heads or tails is true Heads ∨ Tails

Either heads or tails is False ¬Heads ∨ ¬Tails

I win if heads is true Heads → IWin ¬Heads ∨ IWin

you lose if tails is true Tails → YouLose ¬Tails ∨ YouLose

if you lose, I win YouLose → IWin ¬YouLose ∨ IWin

If I win, you lose IWin → YouLose ¬IWin ∨ YouLose

3.2.2 Gee ‘n bewys van “IWin” deur met oplossing (resolution) die teendeel te probeer bewys.
Begin met die negatief van dit wat jy wil bewys en gebruik dan oplossing om ‘n teenstrydigheid (⊥ ) af
te lei met stellings uit die KB.

Show a resolution refutation proof for “IWin” given your KB. Start with the negation of what’s to be
proved, add statements from your KB and use resolution to derive a contradiction. (⊥ ) (5)

Answer:
1 assume ¬IWin
2 given ¬Heads ∨ IWin
3 Resolve 2, 3 ¬Heads
4 given Heads ∨ Tails
5 Resolve 3, 4 Tails
6 given ¬Tails ∨ YouLose
7 Resolve 5, 6 YouLose
8 given ¬YouLose ∨ IWin
9 Resolve 7, 8 IWin
10 Resolve 1, 9 ⊥

Vraag 4. / Question 4. (SU4) [7]

Stel die volgende Engelse sinne voor in predikaatlogika.


Represent the following English sentences in predicate logic
4.1 Every boy is a human who is male and whose age is less than 16 and every male human
who is less than 16 is a boy (3)
4.2 Some people like every vegetable (2)
4.3 There is no vegetable that is liked by every person (2)
Answers (variations are possible, but syntax and semantics are important):
4.1 ∀x boy(x) ↔ human(x) ∧ male(x) ∧ age(x) < 16
4.2 ∃x ∀y person(x) ∧ vegetable(y) → likes(x,y)
4.3 Either is acceptable:
¬ ∃x ∀y vegetable(x) ∧ person(y) → likes(x,y)
∀x ∃y vegetable(x) person(y) → ¬ likes(x,y)
CMPG313 7
Vraag 5. / Question 5. (SU5) [6]

Skryf Afrikaanse sinne waarvanaf elkeen van die volgende wasige logika stellings opgestel sou kon word.
Write English sentences from which the the following fuzzy logic statements could have been constructed:
5.1 Weather_is_nice = 0.7 (1)
5.2 Weather_is_nice = 0.2 (1)
5.3 Coffee_is_Hot = 0.8 (1)
5.4 Coffee_is_Hot = 0.3 (1)
5.5 Skryf u eie wasige logika (fuzzy logic) sin, gevolg deur die vertaling daarvan in Afrikaans.
Write your own fuzzy logic sentence, followed by its translation in English (2)
Answers:
5.1 The weather is fairly nice
5.2 The weather is not very nice/ terrible
5.3 The coffee is extremely hot/ considerably hot
5.4 The coffee is relatively cold
5.5 Depends on the student's answers

Vraag 6. / Question 6. (SU7+SU8) [12]


Vergelyk diepte-eerste soek, breedte-eerste soek en beste-
eerste soek op die volgende grafiek, deur die onderstaande
vrae te beantwoord.
Die beginnode is D en die doelwitnode is M.
Die syfers dui afstand tussen nodes aan vir gebruik deur die
beste-eerste algoritme.

Compare depth-first search, breadth-first search and best-


first search on the following graph by answering the
questions below.
The start node is D and the goal node is M.
The numbers show the distance between nodes for use by
the best-first algorithm.

6.1 Vir diepte-eerste en breedte-eerste, gee die volgorde


van nodes wat geëvalueer word van I tot J. (2)
For depth-first and breadth-first, give the sequence in which the nodes are evaluated from I to J.
Mistake in question paper. It should be ... “from D to M”

6.2 Vir beste-eerste, toon al jou stappe duidelik aan en wys by elke stap die volgende (bv. in 'n tabel):
Die node wat geëvalueer word, die oop lys en die geslote lys. (6)
For best-first, show all your steps clearly and at each step show the following (e.g. in a table):
The node being evaluated, the open list, the closed list.

6.3 Watter een van hierdie algoritmes evalueer die minste nodes? Sê ook hoeveel dit evalueer. (2)
Which one of these algorithms evaluates the least nodes? Also, specify how many it evaluates.

6.4 Watter een van hierdie algoritmes vind die kortste pad? Wat is die kortste pad en sy lengte? (2)
Which one of these algorithms finds the shortest path? What are the shortest path and its length?

CMPG313 8
Question 6 - Answers

6.1 The question should have read “... evaluated from D to M” according to the question introduction.
Depth-first: D, E, H, K, G, J, M
Breadth-first: D, E, F, I, H, K, A, B, C, G, J, M (2)

NOTE: Due to the mistake in the question paper at 6.1


any attempt to answer 6.1 will be credited with 2 marks.

6.2 Best-first steps (12 rows x 0.5 = 6)

Evaluate Open (Important: sorted by total distance) Closed notes/comments


D
D F1, I1, E3 See Note 1 below
F I1, E3 D
I B2, E3, A3, C5 D,F
B E3, A3, L3, C5, M6 D,F,I See Note 2 below
E A3, L3, K4, C5, M6, H6, D,F,I,B
A L3, K4, C5, M6, H6 D,F,I,B,E
L K4, C5, M6, H6 D,F,I,B,E,A
K C5, J5, M6, H6, G8 D,F,I,B,E,A,L
C J5, M6, H6, G8 D,F,I,B,E,A,L,K
J M6, H6, G8 D,F,I,B,E,A,L,K,C
M Found M (path length 6) D,F,I,B,E,A,L,K,C,J

Note 1: Always add child nodes from left to right, but sorted according to total path distance.

Note 2: Although M (goal) is encountered here as a neighbour of B, it must first be added to the Open
list (sorted by total distance) because the current path may not be the shortest path to the goal. When
M is at the front of the sorted open list, then the path will be shortest.

6.3 Depth-first
It evaluates 7 nodes (2)

6.4 Best-first finds the shortest path


path: D – I – B – M of total length 6 (2)

CMPG313 9
Vraag 7. / Question 7. (SU8) [ 7]

Beskou die speletjieboom hieronder waarin die eerste speler haar telling probeer maksimeer.
Die getalle aan die blare is die waardes wat 'n statiese beoordelaar vir die bordposisies bereken het.
Consider the game tree below in which the first player is trying to maximize her score and the number
at the leaves are the values returned by a static evaluator for the board positions reached.
7.1 Vul elke blokkie in met die waarde wat deur die standaard minimaks algoritme teruggestuur word
Fill in each box with the value returned by the standard minimax algorithm (6)

7.2 Omkring die beste aanvanklike skuif vir die eerste speler: A / B / C
Circle the best initial move for the first player: A / B / C (1)
Answer: (Look at red values below. Minus ½ for each wrong value.)

7.2 Circle (C). C is the best initial move for the first player
Note:
the answer must indicate the choice of “C” and not just give the values for A, B and C

CMPG313 10

You might also like