JOURNAL OF COMPUTER SCIENCE AND ENGINEERING, VOLUME 16, ISSUE 1, NOVEMBER 2012

1

Revised General Test/Gross Point Average
System via Fuzzy Logic Techniques
Tarik A. Rashid and Haval A. Ahmed
Abstract During the last decade, the idea of the theory of fuzzy sets has been more resolutely established as an innovative method to
handle the theory of decision processes and complex systems. Fuzzy logic is a way to solve problems through calculation based on degree
of truth as oppose to the traditional way of true or false. This paper aims at modeling and designing a student administration system based on
revised general test/gross point average in the registration department in the third level educational system via fuzzy logic techniques. Two
models were designed for this purpose. These models will take three input variables namely; revised general test, gross point average, rank,
then the models will produce an output either admit or not admit. The accuracy of the models was found to be very encouraging.

Index TermsStudents admiration system, Revised general test, Gross point average, Fuzzy inference system

1 INTRODUCTION

UZZY logic idea was first progressed by Lotfi
Zadeh of the University of California at Berkeley
in the 1960s [1, 2, 3, 4, 5]. This idea was revolu-
tionized when Dr. Zadeh was manuplating natural
language and in particular with computer problem to
understand natural language [1, 2, 3, 4, 5]. Natural
language is the most common activity in our daily
life, in fact representing this problem into the abso-
lute zeros and ones is an uphill task [1, 2, 3, 4,
and 5].

Right from 1990s, the applications of fuzzy logic
have been vividly augmented and improved to cover
finance, marketing, weather forecasting, manufactur-
ing, academic research, controller based systems,
Pharmaceutics and many more decision making
problems [2, 6, 7, 8, 9].

Fuzzy logic techniques have emerged as some of
the best and most promising approaches to deal
with problems and systems that involve nonlineari-
ties. Fuzzy logic techniques are known for their
abilities to manipulate the problems that are complex
to solve nonlinear and vaguely defined by integrating
a preexisting knowledge and expertise of the system
and behavior [2, 10].

Fuzzy logic systems are vital and crucial to deal
with problems that are mathematically weak or when
the mathematical models are poorly defined to gain
much more apparent and considered qualitative per-
ception and understanding.

Wang [10, 2] mentioned that nonlinear functions
can be generated by fuzzy control theory according
to his representation theorem. Wang specified that
any continuous non-linear function can be represent-
ed as exactly as needed with a finite set of fuzzy
variables, values and rules. Thus, it is commonly
possible to develop a fuzzy control system in which
it is appropriate for non-liner systems using suitable
design and procedures [10, 2].

In this paper, fuzzy logic techniques are used to
solve the problem of a student admission through
determining his or her Revised General Test/Gross
Point Average System and ranking.

The paper is organised as follows: in section 2,
we describe the basic definitions of fuzzy set theory,
fuzzy membership functions, fuzzy linguistic rules
and defuzzyfications. In section 3, fuzzy develop-
ment models and rule forming are explained. In
Section 4, simulations and results are explained;
finally, at the end, the main conclusions are out-
lined.
2 BASIC DEFINITIONS OF FUZZY SET
THEORY
The study of fuzzy set involves different concepts;
these are: fuzzy sets and membership functions,
fuzzy linguistics and defuzzyfication. The subsequent
sub-sections are brief explanations for the aforemen-
tioned concepts.

2.1 Fuzzy Sets and Concept of Membership
Consider A as a standard set; in other words, A
F
2012 JCSE
www.Journalcse.co.uk

- Tarik A. Rashid is with the Software Engineering, College of
Engineering, Salahaddin University, Hawler, and Kurdistan.
- Haval A. Ahmed is with the Sfotware Engineering, College of
Engineering, Salahaddin University, Hawler, and Kurdistan.


2



can be defined as a collection of elements or ob-
jects, and any arbitrary element or object x either
belongs or does not belong to A. The membership
) (x
A

of x in A can be expressed as followings

[1, 2]:-

} 1 , 0 { ) ( = X x
A


That is, it may take the value 1 or 0, which rep-
resent the truth value of x in A. Let the compliment
of A be given as A and the symbol

be given
as intersection between the two sets, then the inter-
section of the A and A is not null as it is ex-
pressed below:

| = ' A A

The idea of fuzzy logic is actually a logic based
on fuzzy sets; for example, sets of elements or ob-
jects described by truth-values in the [0, 1] interval
rather than crisp 0 and 1, as in the conventional set
theory. The membership function can be described
as the function that allocates a number in [0, 1] to
each element of the universe of discourse of a
fuzzy set [1, 2, 3, 4, 5].
2.2 Membership Functions
Consider that X signifies the space of discourse of
a fuzzy set A.

| | 1 , 0 ) ( = X x
A


If fuzzy sets are selected, then a membership
function for each and every set should be generat-
ed. A membership function is a distinctive curvature
that transfigures the numerical value of input within
a range from 0 to 1, signifying the belongingness of
the input to a fuzzy set [1, 2, 3, 4, 5]. This is
called fuzzification. There are different forms of
membership function, such as triangle, trapezoid,
Gaussian, etc. The very basic and simplest form of
these is a Triangular membership function. It can be
represented by a collection of three points that form
a triangle [11, 12, 13, 14] defined as a triangular
membership function, as can be seen in Fig. 1.
The triangular membership function can be ex-
pressed mathematically as following:-



The second type of membership function is called
a trapezoidal membership function as shown in Fig.
2.




The trapezoidal membership function can be ex-
pressed mathematically as follows:-




























Fig. 2. displays trapezoidal membership function.






Fig. 1. displays a Triangular membership function.
3



The linear decreased membership function is
seen clearly in Fig. 3.



The linear decreasing membership function can
be expressed as follows:-











The linear increased membership function is seen
Fig. 4.



The linear increasing membership function can be
expressed as follows:-

The Gaussian membership function can be seen
in Fig. 5.



Gaussian function can be expressed as follows:



The Gaussian membership function depends on
two parameters, standard deviation () and means
(G), and it is represented as shown above.

2.3 Fuzzy Linguistic Rules
It is clear that quantitative reasoning can be gener-
ated by fuzzy linguistic rules that relate input fuzzy
sets with output fuzzy sets. In fact, the fuzzy-rule
base comprises a number of fuzzy if-then rules. For
example, if we consider that we can have two-inputs
and one output fuzzy system; this can be represent-
ed as follows:

Rule: If x is medium and y is low then z is high.















Fig. 3 displays the linear decreased function.









































Fig. 4 displays the linear increased function.















Fig. 5. Displays the Gaussian function.
4




Where x, y and z are variables representing two
inputs and one output; medium, low and high are
the fuzzy sets of x, y and z, respectively.
2.4 Defuzzifications
It is considered that the output of each rule is also
a fuzzy set. All the output fuzzy sets are then
gathered and collected into one single fuzzy set.
This process is called aggregation [1, 2, 3, 4, and
5]. In the end, the output set is determined to a
single crisp number by defuzzification. Various de-
fuzzification methods are commonly used in the lit-
erature such as centroid, center of sums, mean of
maxima and left-right maxima. It is clear that the
centroid method of defuzzification is widely used in
most of the cases [1, 2, 3, 4, 5, 17]. It is done as
shown below:







Where x
def
is the defuzzified output and (x) is
the output fuzzy set after aggregation of individual
implication results.

2.5 Fuzzy set operations
In order to design fuzzy models, we need to have
knowledge and understanding of the operations of
the theory of fuzzy sets. The fuzzy set operations
are defined with respect to the sets' membership
functions.

Two fuzzy sets A and B on the universe of dis-
course X are equal if their membership functions
are equal for each x
e
X:


) ( ) ( : x x X x
B A
= e

A fuzzy set A is a subset of B (A _ B) if


) ( ) ( : x x X x
B A
s e


For the operation of intersection

of two fuzzy
sets A and B, there are different forms of definition
in the literature. The choice is application dependent
[1, 2]:


The union

of two fuzzy sets A and B is also

defined in several ways:

3 DEVELOPMENT OF MODEL
3.1 Input variables
A fuzzy inference system has been built to imple-
ment an application of Revised General Test/Gross
Point Average model. The model is fed with three
parameters; these input variables are Revised Gen-
eral Test (GRE), Gross Point Average (GPA) and
Rank. The model has one output only admit or non-
admit. The first input GPA was divided into three
edges; there are Low, Medium, and High. See Fig.
6.



















Fig. 6: displays the mapping of GPA input variable.














Fig. 6: displays the mapping of GPA input variable.














Fig. 6: displays the mapping of GPA input variable.
5










GRE is the second input variable from which three
edges are formed namely; Low, Medium, High, and
Very High. See Fig. 7.

The third variable is divided into three edges
namely; Low, Medium, High. See Fig. 8

The output variable is divided into four edges
namely; Low, Medium, High and very high, see Fig.
9.

3. 2 Forming rules
The math lab software is used to implement
GRE/GPA test model. Two different models are
used. The first model contained 7 rules; see Table
1. The second model contained 19 rules; see Table
2, the rules were selected carefully.


















Fig. 8: displays the mapping of Rank input variable.















Fig. 9: displays the mappings of output variable admit.














Fig. 7: displays the mapping of GRE input variable.
Rules GPA GRE Rank Admit/
Non-
Admit
1 Low Low Low Poor
2 Med. Med. Med. Med.
3 Med. V. High High V. High
4 High High High V. High
5 High High V.High
6 High V. High V.High
7 High V. High High V.High
TABLE 1
THE FIRST MODEL
6




4 RESULTS
In this paper, math lab software is used to develop
both models, the first model is used with 7 rules
and the second model is used with 19 models. Fig.
10 shows the results of the first model.

Fig. 11 shows the surface for the first model. Fig.
12 shows the result for the second model. Based
on the performance criteria in [18], it is clear that
the second model with 19 rules outperformed the
first model. Both models are submitted to the same
input and output parameters along with the same
membership functions. It is clear the second model
produced a better result than the first model. The
max error value for the first mode was 0.459,
whereas the max error for second model was 0.370
and the reason for this accuracy was due to the
formation of more accurate fuzzy rules in the sec-
ond model. Bear in mind the max error will increase
as we increase the number of testing samples. Fig
13 shows the surface for the second model.
























TABLE 2
THE SECOND MODEL
Rules GPA GRE Rank
Admit/
Non-Admit
1 Low Low Low Poor
2 Low Med. Low Poor
3 Low Med. Poor
4 Low High Med.
5 Low V.High High
6 Med. Low Poor
7 Med. Med. Med. Med.
8 Med. High High High
9 Med. High High
10 Med. V. High V.High
11 Med. V. High High V.High
12 High Low low Med.
13 High Low Med.
14 High Med. Med. High
15 High Med. High
16 High High High V.High
17 High High V.High
18 High V.High V.High
19 High V.High High V.High


.









Figure 10: displays the results of the first model.





.







Figure 11 shows the surface of the first model.



7





5 CONCLUSION
In this paper many general ideas of the basics of
fuzzy reasoning have been introduced. We also
have introduced the theory of fuzzy sets, along with
some important definitions related to the member-
ship function, logical and transformation operators,
fuzzy relations, implication and inference rules, and
fuzzy similarities and differences measures have
been stated.

The fundamental structures of a fuzzy inference
system and its elements have been explained.

The reason behind using fuzzy logic techniques is
based their ability in solving complex, nonlinear
and/or ill-defined problems over more conventional
approaches, which lies in their capability to incorpo-
rate a priori qualitative knowledge and expertise
about system behavior and dynamics.

In this paper an application of Revised General
Test/Gross Point Average System via Fuzzy Logic
Techniques is presented. Three input variables were
taken as inputs to the two models of fuzzy inference
systems. The two systems have one output. Under-
standing the application is very important to develop
these two models and extracting the fuzzy rules.

The first model is developed with 7 rules where
the second model is developed with 19 rules. Both
models are submitted to the same input and output
parameters, along with the same membership func-
tions. It is clear the second model produced a better
result than the first model. The max error value for
the first mode was 0.459, whereas the max error for
second model was 0.370, and the reason for this
accuracy was due to the formation of more accurate
fuzzy rules in the second model.

ACKNOWLEDGMENT
The authors wish to thank Ms. Karen Nilson from
the College of the Marshall Islands and Dr. Rana
Elfara from Science Department, Shuwaifat School,
Hawler, Kurdistan.
REFERENCES
[1] Abhijit Majumdar, Ph.D., Anindya Ghosh, Ph.D,
2Yarn Strength Modelling Using Fuzzy Expert System,
Journal of Engineered Fibers and Fabrics
http://www.jeffjournal.org Volume 3, Issue 4, 2008.
[2] Costas P. Pappis FUZZY REASONING University of Piraeus
Piraeus, Greece Constantinos I. Siettos National Technical
University of Athens School of Applied Mathematics and Physics
Athens, Greece,xxx.
[3] Zadeh, L. A., Fuzzy Sets, Information and Control, 8, 338-
353,1965.
[4] Zadeh, L. A., Outline of a new approach to the analysis com-
plex systems and decision processes, IEEE Trans. Syst. Man.
Cybernet. 3:28-44, 1973.
[5] Majumdar, A., Majumdar, P. K., and Sarkar, B., Application of
an Adaptive Neuro-Fuzzy System for the Prediction of Cotton
Yarn Strength from HVI Fibre Properties, J. Text. Inst., 96,
(2005) 55-60. Journal of Engineered Fibers and Fabrics
http://www.jeffjournal.org Volume 3, Issue 4, 68, 2008.








Fig 13 shows the surface of the second model.


.










Figure 12: displays the results of the second model.



8



[6] Sugeno, M. and Yasukawa, T., A fuzzy-logic-based approach to
qualitative modeling, IEEE Trans. Fuzzy Syst. 1:7-31, 1993.
[7] Sugeno, M. and Kang, G. T., Structure identification of fuzzy
model. Fuzzy Sets Syst. 28:15-23. 1988.
[8] Karr, C. L. and Gentry, E. J., , Fuzzy control of pH using ge-
netic algorithms, IEEE Trans. Fuzzy Syst. 1:46-53, 1993.
[9] Lee, C. C, Fuzzy logic in control systems: fuzzy logic control-
lersparts I and II, IEEE Trans. Syst. Man. Cybernet. 20:404-
435, 1990.
[10] Wang, L. X, Fuzzy systems are universal approximators, in:
Proc. IEEE Int. Conf. on Fuzzy Systems (San Diego), pp. 1163-
1170, 1992.
[11] Dubois, D. and Prade, H., Fuzzy Sets and Systems: Theory
and Applications. Academic, NewYork, 1980.
[12] Dubois, D., and Prade, H., Fuzzy Real Algebra, Some Results,
Fuzzy Sets and Systems, 2, 327-348, 1979.
[13] Zimmermann, H. J., Fuzzy Set Theory and its Applications, 3rd
edn, Kluwer, Dordrecht, 1996.
[14] Zimmermann, H. J., Ruan D. and Huang C, eds, Fuzzy Sets
and Operations Research for Decision Support: Key Selected
Papers, Beijing Normal University Press, Beijing, 2000.
[15] Pappis, C. P. and Mamdani, E. H., A fuzzy logic controller for
a traffic junction, IEEE Systems, Man Cybernet. (SMC-7)
10:707-717, 1977.
[16] Pappis, C. P. and Sugeno, M., Fuzzy relational equations and
the inverse problem. Fuzzy Sets Syst. 15:79-90, 1985.
[17] T. Rashid, Fuzzy Logic and Neuro Fuzzy Models for Direct
Current Motors, International Journal of Engineering Inventions
ISSN: 2278-7461, www.ijeijournal.com ,Volume 1, Issue 7, PP:
68-75, 2012.
[18] T. Rashid, T. Kechadi, B. Gleeson, Artificial Intelligence Based
Methods for Electric Load Forecasting, the third international
confernce on cybernetics and inforamtion technologies systems
and applications (CITS 2006) in Orlando USA, 2006.
[19] Berkan, R. C., and Trubatch, S. L., Fuzzy Systems Design
Principles, Standard Publishers Distributors, New Delhi, (2000)
22131, 2000.
[20] Kartalopoulos, S. V., Understanding Neural Networks and Fuzzy
Logic: Basic Concepts and Applications, Prentice-Hall of India
Pvt. Ltd., New Delhi, (2000)75-82, 2000.
[21] Klir, G. J., and Yuan, B., Fuzzy sets and Fuzzy logic: Theory
and Applications, Prentice-Hall of India Pvt. Ltd., New Delhi,
(2000) 35-49, 2000.
[22] Bector, C. R., and Chandra, S., Fuzzy mathematical program-
ming and Fuzzy Matrix Games, Springer, Berlin, (2004) 21-56,
2004.
[23] Klein, W., The Technology of Short Staple Spinning, Volume
1, The Textile Institute, (1987) 2-9. L.A. Zadeh, Fuzzy Sets
And Applications , Ed. R.R. Yager Et Al. (John Wiley, New
York, 1987)"U.S., 1987.
[24] Steven D. Kaehler, Fuzzy Logic An Introduction Part 1 And
2,
Http://Www.Seattlerobotics.Org/Encoder/Mar98/Fuz/Fl_Part1.
Html# Introduction, 1998.
[25] Quek, C., & Zhou, R. W., The Pop Learning Algorithms: Re-
ducing Work In Identifying Fuzzy Rules, Neural Networks,
14(10), 1431-1445, 2001.
[26] X.XXX Europe Gets Into Fuzzy Logic, (Electronics Engineering
Times, Nov. 11, 1991).1991.
[27] Loses Focus On Fuzzy Logic" Machine Design, June 21, 1990.
[28] By Emily T. Smith, Why The Japanese Are Going In For This
'Fuzzy Logic. Business Week, Feb. 20, 1993.
[29] Abraham A., "Adaptation Of Fuzzy Inference System Using
Neural Learning, Fuzzy System Engineering: Theory And Prac-
tice", Nadia Nedjah Et Al. (Eds.), Studies In Fuzziness And Soft
Computing, Springer Verlag Germany, ISBN 3-540-25322-X,
Chapter 3, Pp. 5383, 2005.
[30] Tharwat E. Alhanafy, Fareed Zaghlool And Abdou Saad El
Din Moustafa, Neuro Fuzzy Modeling Scheme For The Predic-
tion Of Air Pollution, Journal Of American Science, 6(12) 2010.
[31] Adriano Cruz, ANFIS: Adaptive Neuro-Fuzzy Inference Sys-
tems, Mestrado NCE, IM, UFRJ, Logica Nebulosa P. 1/33.
[32] T. M. Nazmy, H. El-Messiry, B. Al-Bokhity, Adaptive Neuro-
Fuzzy Inference System For Classification Of Ecg Signals,
Journal Of Theoretical And Applied Information Technology, Pp-
71-76, 2010.
[33] R. Rojas: Neural Networks, Fuzzy Logic, Springer-Verlag, Ber-
lin, 1996.
[34] Jyh-Shing And Roger Jang., Anfis: Adaptivenetwork-Based
Fuzzy Inference System, Computer Methods And Programs In
Biomedicine, Ieee Transactions On Systems, University Of Cali-
fornia, 1993.
[35] Abdulkadir Sengur., An Expert System Based On Linear Dis-
criminant Analysis And Adaptive Neurofuzzy Inference System
To Diagnosis Heart Valve Diseases, Expert Systems With Appli-
cations, 2008.
[36] G. Zhao, C. Peng And Xiting Wang., Intelligent Control For
Amt Based On Drivers Intention And Anfis Decision-Making,
World Congress On Intelligent Control And Automation, 2008.
[37] Anupam Das, J. Maiti And R.N. Banerjee., Process Control
Strategies For A Steel Making Furnace Using Ann With Bayesi-
an Regularization And Anfis, Expert Systems With Applications,
2009.
[38] Y. Jin, Fuzzy Modeling Of High-Dimensional Systems: Com-
plexity Reduction And Interpretability Improvement, Ieee Trans-
actions On Fuzzy Systems, 8(2), 212-221, 2000.
[39] Zhou, R. W., & Quek, C., Popfnn: A Pseudo Outer-Product
Based Fuzzy Neural Network, Neural Networks, 9(9), 1569-
1581.1996
9



[40] Quek, C., & Zhou, R. W. , Popfnn-Aar(S): A Pseudo Outer-
Product Based Fuzzy Neural Network, Ieee Transactions On
Systems, Man And Cybernetics, Part B, 29(6), 859-870, 1999.
[41] Ang, K. K., Quek, C., & Pasquier, M., Popfnn-Cri(S): Pseudo
Outer Product Based Fuzzy Neural Network Using The Compo-
sitional Rule Of Inference And Singleton Fuzzifier, Ieee Trans-
actions On Systems, Man And Cybernetics, Part B, 33(6), 838-
849, 2003.
[42] Ang, K. K., & Quek, C., Rspop: Rough Set-Based Pseudo
Outer-Product Fuzzy Rule Identification Algorithm, Neural Com-
putation, 17(1), 205-243,2005.
[43] Kosko, Bart, Neural Networks And Fuzzy Systems: A Dynam-
ical Systems Approach To Machine Intelligence, Englewood
Cliffs, Nj: Prentice Hall. Isbn 0-13-611435-0, 1992.
[44] Lin, C.-T., & Lee, C. S. G., Neural Fuzzy Systems: A Neuro-
Fuzzy Synergism To Intelligent Systems, Upper Saddle River,
Nj: Prentice Hall, 1996.
[45] A. Bastian, J. Gass, Selection Of Input Variables For Model
Identification Of Static Nonlinear Systems", J. Of Intelligent And
Robotic Systems, Vol. 16, Pp.185-207, 1996.

Tarik A. Rashid received the Bachelor Science Degree in Engineer-
ing from Mosul University in 1990 and his Post Graduate Degree in
computing from Griffith College Dublin (GCD) in 2000-2001, Dublin-
Ireland.

He received his Philosophy Doctorate Degree in Computer Science
and Informatics, College of Engineering, Mathematical and Physical
Sciences from University College Dublin (UCD), 2001-2006, Dublin-
Ireland.

Dr Tarik Rashid was a postdoctoral fellow and senior member of
Parallel Computational Research Group (PRCG) at University Col-
lege Dublin 2006-2007, Dublin-Ireland.

He was a director academic affair at the College of Computer Train-
ing (CCT) 2007-2009, Dublin, Ireland.

Tarik was a dean of college of Information Technolgoy, Business and
Management Studies at SABIS University, 2009-2010, Hawler, and
Kurdistan.

Tarik is currently an assistant professor at Software Engineering
Department, College of Engineering, Salahaddin University, 2010-
present, Hawler, Kurdistan.

Haval A. Ahmed is currently a Master degree student in Software
Engineering Department, College of Engineering, Salahaddin Uni-
versity, 2010-present, Hawler, Kurdistan.