Simulation and Neural Networks

haremboingΤεχνίτη Νοημοσύνη και Ρομποτική

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

61 εμφανίσεις

ميحرلا نمحرلا الله مسب

College of Information Technology


Course
no

:

CSCI 5
5
2

MS in CS & IST

Student No: __________

Course Title:

Current

Final

Exam

Student Name:

Development in Computer
Science I (Neural Networks)

Theoretical Exam

___________________
__


[Page
1

of
2

Pages
]



Exam
Time:
3
:
0
0
Hours

3
rd



quadmester
2005/2006

Date:
8

/

6

/200
6

3
:
3
0
PM

No of Questions
:
F
ive

Total Grade: 100

Try to Answer
All

Questions

Open Book:
Yes


Using Computer:
No

Using Calculator:
Yes




Attempt to Solve F
ive

Questions

First Question


2
0
/100


Simulation and Neural Network
s

a)

Compare between the training
styles with

static NN and dynamic NN in
Mat

lab
, Which training style of them is
the best
for classification when
using Kohonen SOM NN Model

and why
.


b)

There are many
Programming languages
, tools and applications integrated
with neural networks
..

discuss that and show when will programming
language be the best and when
will
tools and applications be the best

to be
integrated with neural network
.



Second Question


2
0
/100


Image

Processing

and Neural Network
s

a) There are
many key stages in digital image processing
..

define them briefly

..
When will you need to use Neural network Models

in image processing
..
and with which
stage

… Discuss that with an example..



b)
Is Neural network is

good for face recognition, signature verification and
hand
written

recognition discuss that with different NN models examples.



Third Question


2
0
/100



Control

and Neural Network

a)

When using Neural network for controlling system which is more suitable
the single layer perceptron or the multi layer and why.


b)
Neural Network is a good model for controlling processes Discuss that with
two different example
s

where the first
one
by using

a hybrid neural
networks

and the second one by using

the hieratical

neural networks





ميحرلا نمحرلا الله مسب

College of Information Technology


Course
no

:

CSCI 5
5
2

MS in CS & IST

Student No: __________

Course Title:

Current

Final

Exam

Student Name:

Development in Computer
Science I (Neural Networks)

Theoretical Exam

___________________
__


[Page
2

of
2

Pages
]




Fourth

Question


2
0
/100


Data Mining and Neural Network
s

a) There are different data mining procedures
..

define
each one
by example
and then discuss it with Neural Networks models.

b)
There are different data mining techniques
define each one briefly and try
to propose
a new hybrid technique to solve
classifying
problems in data
mining.


Fifth

Question


2
0
/100


Speech Processing and Neural Network
s

a) There are three speech recognition Approaches .. When will you need to use

Neural network Models for speech recognition .. and with which approach
… Discuss that with an example..



b) According to kind of speech there are two kinds of speech recognition
classification …For each one propose a neural network model or a Hybrid
NN or a hierarchal NN that is more suitable for that kind.


Sixth

Question


2
0
/100


Answer Four of the following:


a)


How many learning methods
are there
for NN

s ?
Discuss that

?

b)

What can you do with a Neural Network and what
can
not ?

c)

In NN’s

structure e
ach
Neuron

is presented by processor element .. Are we
need memory or
Databases with NN

s

structure … Discuss that..


d)

Compare between the two hybrid systems Genetic Algorithm with Neural
Networks and Fuzzy logic with Neural Networks

which i
s the best and why?

e)

Discuss the future of Neural Networks from your view ..

f)

Ar e Ne ural Ne t wor ks mode l s good f or knowl e dge r epr e s e nt at i on .. why
..e xpl ai n t hat br i ef l y.



End o f Que s t i o ns

Go o d Luc k