Autumn 2006/2007 - The University of Nottingham, Malaysia Campus

roomycankerblossomΤεχνίτη Νοημοσύνη και Ρομποτική

23 Οκτ 2013 (πριν από 3 χρόνια και 7 μήνες)

125 εμφανίσεις

G5BAIM
-
E1

G5BAIM
-
E1


Turn
o
ver



The University of Nottingham

Malaysia Campus



SCHOOL OF COMPUTER SCIENCE AND INFORMATION TECHNOLOGY



A LEVEL B MODULE, AUTUMN SEMESTER 2006
-
2007



ARTIFICIAL INTELLIGENCE METHODS



Time allowed TWO Hours






Candidates must NOT start writing their ans
wers until told to do so



Answer FOUR out of SIX questions


Only silent, self
-
contained calculators with a single
-
line display are permitted in this
examination.


Dictionaries are not allowed with one exception. Those whose first language is not English

may use a standard translation dictionary to translate between that language and English
provided that neither language is the subject of this examination. Subject specific translation

dictionaries are not permitted.


No electronic devices capable of sto
ring and retrieving text, including electronic dictionaries,
may be used.


DO NOT turn examination paper over until instructed to do so


























G5BAIM
-
E1

G5BAIM
-
E1



Turn over

2



1.

This question is concerned with genetic algorithms. (25 marks)


(a)

Describe the main categories

of the Evolutionary Algorithm’s applications.

[9 marks]


(b)

Why is mutation important with regard to genetic algorithms? Give an example
to demonstrate your point.








[3
marks]


(c)

Explain the difference between a genotypic representation and a phenotypic
r
epresentation. Give an example of each.





[8
marks]


(d)

Outline the similarities and differences between genetic algorithms and genetic
programming.









[5
marks]



2.

This question is concerned with genetic algorithms. (25 marks)


(a)

What do you understand b
y the following Evolutionary Strategies
notations:

[3 marks]


i)

(1 + 1) ES


ii)

(

+

) ES


iii)

(

,

)

ES


(b)

Given the following parents,
P
1

and P
2
, and the template T as shown:


P
1

A

B

C

D

E

F

G

H

I

J

P
2

E

F

J

H

B

C

I

A

D

G

T

1

0

1

1

0

0

0

1

0

1


Assume that
C
n
’s are crossover points where C
n

= 0 means that the crossover
point is at the extreme left of the parent. Show how the following crossover
operators work with regard to genetic algorithms.






i)

one point crossover (using C
1

= 4)





[4 marks]


ii)

two point crossover (using C
2

= 2 and C
3

= 8)


[4 marks]


iii)

uniform crossover (consider the Mask T)



[4
marks]



iv)

order
-
based crossover (C
4

= 2) (ignore T)


[3 marks
]


(c)

Give an example that shows how point based crossover may fail (produce an
unacceptable solution) and how a Partially Mapped Crossover (PMX) or an order
crossover might solve the anomaly.






[7
marks]



G5BAIM
-
E1

G5BAIM
-
E1



Turn over

3




G5BAIM
-
E1

G5BAIM
-
E1



Turn over

4



3.

This question is concerned with evoluti
onary programming/strategies. (25 marks)


(a)

Outline the similarities and differences between genetic algorithms and
evolutionary strategies.








[5
marks]


(b)

Explain “Self
-
Adaptive Behaviour” in Evolutionary Strategies.


[5
marks]


(c)

Suggest a rea
l
-
world problem and describe how it can be solved using GA. The
discussions should cover the points listed below:


i)

The nature of the problem and its complexity.



[3 marks]


ii)

How GA can be used appropriately to deal with the complexity,
in relation
to the points listed below:




1.

How the representation is defined (phenotype, genotype)?


[1.5 marks]


2.

What are candidate solutions (population)?

[1.5 marks]


3.

How to evaluate each population? [1.5 marks]


4.

What is the fitness value due to the objective function? [1.5 marks]


5.

How members can survive and be omitted (selection)? [1.5
marks]


6.

How

the new individual is generated (crossover, mutation)?


[1.5 marks]


iii)

The termination condition and related discussions.





[3 marks]



4.

This question is concerned with semantic nets, tabu search and hill
-
climbing. (25
marks)


(a)

Outline the significant features of semantic nets as defined by Quillian, and
explain Simon’s solution for the inefficiency of the early semantic
nets.


[6
marks]


(b)

List down the set of four primitive conceptualizations for building the world of
meaning as introduced by Roger Shank.





[2
marks]


(c)

Represent the sentence “John will welcome guests” using conceptual
dependency theory and notations.








[5 marks]


(d)

Write brief notes on tabu search and state on which aspect of the real world it is
based.










[6
marks]


(e)

What is the main difference(s) between simulated annealing and hill climbing?

G5BAIM
-
E1

G5BAIM
-
E1



Turn over

5

[6 marks]






G5BAIM
-
E1

G5BAIM
-
E1




End

6



5.

This question is concerned with simu
lated annealing. (25 marks)


(a)

Show a simulated annealing algorithm.






[5
marks]


(b)

Outline the simulated annealing cooling schedule, describing the various
components.










[5
marks]


(c)

The following table shows six evaluations of a simulated annealing
a
lgorithm.


i)

For each evaluation give the probability of the next state being accepted.
Assume the objective function is being maximised.



[5
marks]


No.

Current State

(Evaluation)

Potential New State

(Evaluation)

Temperature

1

120

50

20

2

120

50

500

3

120

100

20

4

120

100

500

5

120

150

20

6

120

150

500


ii)

Discuss the results from part (i).






[10
marks]



6.

This question is concerned with ant colony and fuzzy logic. (25 marks)


(a)

Describe how ants are able to find the shortest path to a food sourc
e.

[5
marks]


(b)

Suggest five suitable problems, other than the travelling salesman problem, that
an ant algorithm could be used to solve.





[5
marks]


(c)

Fuzzy logic is a paradigm that can be used in an alternative design methodology
for developing systems fo
r embedded control.


i)

List down the main reasons for using fuzzy logic.



[6
marks]


ii)

Describe the impact of fuzzy logic on design complexity, development
cycle and performance of an embedded control system.


[9
marks]