PREDICTING ACADEMIC MAJOR OF STUDENTS USING BAYESIAN NETWORKS TO THE CASE OF IRAN

International Journal of Computer-Aided Technologies (IJCAx) Vol.2, No.3, July 2015
DOI:10.5121/ijcax.2015.2304 47
PREDICTING ACADEMIC MAJOR OF STUDENTS
USING BAYESIAN NETWORKS TO THE CASE OF
IRAN
Shiva Asadianfam 1
, Mahboubeh Shamsi 2
, Sima Asadianfam 3
1
Department of Computer Engineering, Qom Branch, Islamic Azad University, Qom,
Iran
2
Department of Electrical & Computer Engineering, Qom University of Technology,
Qom, Iran
3
Department of Computer Engineering, Zanjan Branch, Islamic Azad University, Zanjan,
Iran
ABSTRACT
In this study, which took place current year in the city of Maragheh in IRAN. Number of high school
students in the fields of study: mathematics, Experimental Sciences, humanities, vocational, business and
science were studied and compared. The purpose of this research is to predict the academic major of high
school students using Bayesian networks. The effective factors have been used in academic major selection
for the first time as an effective indicator of Bayesian networks. Evaluation of Impacts of indicators on
each other, discretization data and processing them was performed by GeNIe. The proper course would be
advised for students to continue their education.
KEYWORDS
Academic major selection, Field selection, Bayesian networks, GeNIe.
1. INTRODUCTION
Education is one of the major institutions of society are considered. Institutional arrangements
can leave dramatic changes in many areas of social. Analyze the situation of progressive
countries, we find that, they are paying attention to how education in recent decades has led to
progress in various fields [1]. Dynamic and streamlined educational institution, an institution in
which human resources are educate properly and according to the goals. The institutional knows
needs, abilities and talents of people and its prosperity put forward regular programs consistent
with academic standards [2]. Different nations due to of their efficient use of manpower and
resources available study and careful planning. Because scientific research and careful planning
that countries can use them to make the most of their potential in the global
competition. Infrastructure planning for the institution of education is done [2].
In IRAN, to achieve the goals a high school education: "is to prepare students for employment or
further education", there are the short-term and long-term planning, such as budgeting field of
study in education. The new system of secondary education, there are five branch or field of study
for further education. These fields of study are Technical & Vocational, Work & Knowledge,
Mathematics & Physics, Experimental Sciences & Human Sciences.

48
Variety of fields in secondary education and higher education on the one hand, and a wide range
of jobs in today's society and the relationship between education and job on the other will require
students to their future educational and career based on personal interests and talents are correct
planning. Today, education and employment are considered the most important questions in
education and other social and economic aspects.
This paper is structured as follows. In Section 2, the background information required for the
better understanding of the method presented in this paper is discussed. In Section 3, the
empirical studies on Bayesian Networks are presented. In Section 4, material and methods are
explained.Then, implementation of Bayesian Network by genie is presented, and in Section 5,
the results of the implementation are presented. Finally, in Section 6, the general conclusion is
discussed.
2. BACKGROUND
Investigation on education issues and adaptation of results with predefined objectives can show
the failure of the process of education and action to solve .What is evident in the current situation,
the importance of fields students are in high school. High school graduates its first year based on
interest, aptitude and performance in education and training in the assessment and preparation
process changes are guided to one of the following paths: [3]
- Branch Of Theoretical (Mathematics And Physics - Experimental Sciences- Human
Sciences)
- Technical And Vocational Categories
- Branch Of Work And Knowledge
But what on a student's choice of academic major influences, will also be important, select a field
in the near future will bring jobs. In this section, will be discussed the factors influencing choice
of academic major or students' choice of occupation.
2.1. The effective factors in choosing an academic major
Effective factors in the choice of academic major can be divided into two parts:
Individual factors: [4]
- Talent and mental ability
- Willingness and interest
- Personality traits
- Knowledge and skills
- Academic records
- Personal needs
- Sex
- Personal experiences
Environmental factors: [4]
- Cultural Foundations
- Social position
- Geographic Location
- Economic situation
- industry changes
- Religion
- Family and Friends
- School
- Family
- Academic guidance
These are all hand in hand to the person to choose their field of study, but given the subject matter
here is trying that important factors to be considered and given these factors, academic major
must be selected for further study.

49
2.2. Bayesian Networks
If (a1, a2, ... , an) be a set of random variables, Bayesian Networks can express the probability of
any combination of them. Based on Bayesian Networks, Bayesian rule that can be expressed as
(1).
(1)
This network is a natural tool to address two problems in engineering and applied mathematics
that provide uncertainty and complexity of the issues.
A Bayesian Network is called Belief Network or directional probabilities networks, a graph
possibilities show the random variables and their dependencies. The network nodes represent
random variables and arcs represent dependencies between the random variables with the
conditional probability. The network is a graph with directional and no loop. so, All edges are
directional and there are no loops [5, 6].
3. RELATED WORK
There is a wide range of application for Bayesian Network. In the paper [7], a Bayesian Network
model was built to represent the probability distribution of each factor and how they affect
defects, considering strong or weak correlations are existed between individual metric attributes.
The software metrics and build a Bayesian Network model for defect prediction.
Khlifia Jayech and Mohamed Ali Mahjoub propose to use new approach combining distance
tangent, k-means algorithm and Bayesian network for image classification in [8]. First, they use
the technique of tangent distance to calculate several tangent spaces representing the same image.
The objective is to reduce the error in the classification phase. Second, they cut the image in a
whole of blocks. For each block, they compute a vector of descriptors. So, they use K-means to
cluster the low-level features including color and texture information to build a vector of labels
for each image. Finally, they apply five variants of Bayesian networks classifiers to classify the
image of faces using the vector of labels.
In the paper [9] proposed a new framework for discovering interactions between genes based on
multiple expression measurements.This framework builds on the use of Bayesian Networks for
representing statistical dependencies.
In the paper [10], Pablo Felgaer and Paola Britos define an automatic learning method that
optimizes the Bayesian Networks applied to classification, using a hybrid method of learning that
combines the advantages of the induction techniques of the decision trees (TDIDT-C4.5) with
those of the Bayesian Networks. The resulting method is applied to prediction in health domain.
In [11], the aim of study is model development for financial distress prediction of listed
companies in Tehran stocks exchange (TSE) using Bayesian networks (BNs). it can be an

50
evidence that the financial statements of companies have information content. With respect to the
remainder variables in developed models in this research we find firms that have lower
profitability and have more long term liabilities and have lower liquidity are more in risk of
financial distress.
4. MATERIAL AND METHODS
4.1. Random Variables And Data Collection
The study was conducted at Maragheh city In IRAN this year, number of students of secondary
school in the academic major, Technical And Vocational, Work And Knowledge, Mathematics
And Physics, Experimental Sciences And Human Sciences via questionnaire were compared to
the parameters and their values are shown in Table1.
Table 1 -Random variables used in the Bayesian Network
Values
Description
Random variable name
Very much
Much
Low
very low
junior high school grades
High_school_score
School grades
Middle_school_score
Hush threshold score
Not_come_score
Way into university
University
Parental field of study
Paremt_major
Parents Guidelines
Parent_guide
Teachers Guidelines
Teachers_guide
Principal Guidelines
Manager_guide
Consultant Guidelines
Adviser_guide
Friend Recommend
Firend_advise
Program Overview of Introduction jobs
Weeky_plan
field of study relatives
Kinfolk_major
Jobs of the Future
Job
Awareness of community needs
Sociaty_requirment
view than field of study
Sociaty_lookout
The job social status
Social_position
The job salary and wage
Salary
Interest
Tendency
4.2. Form of Bayesian Networks
In this paper, the software GeNIe is used to creat a Bayesian Network. First, identify and show
relationships between random variables in the network. Given the important factors in choice of
academic major with the data obtained from the research that has been done in the city.Also,
consult with an academic advisor are obtained relationships between variables. The frequency and
the percentage effect of each agent is calculated according to Quad ratings in students' choice
academic major for each question and the average effect of each factor according to the number
of students choices each of the four levels range from very low ,low, high, and very high have
been calculated. Finally, after the creation of networks and test them, model is presented, which
consists of 19 nodes and 18 links is shown in Figure 1.
Due to the weight given to each of Quad spectra and using the following formula, the frequency
of each of the variables measured and the mean impact of factor was obtained that the full impact
factors directly connected with a direct link to the target [9].
100
)
(i
f
=
frequency percentage every spectrum of factors (very much, much, low, very low)

51
100
4
3
2
1 )
4
(
)
3
(
)
2
(
)
1
( ×
+
×
+
×
+
× i
i
i
i f
f
f
f
=
average effect of each factor
Figure1- Bayesian network used to predict the field of study
4.3. Chart Of Random Variables
According to the data collected and discretization of data, diagrams of random variables were
drawn that two samples are shown in Figure 2 and 3.
Figure3-Diagram of data frequency of
parent_major
Figure2- Diagram of data frequency of jjob
4.4. Implementation of Bayesian network by GeNIe
Bayesian Network was drawn after identifying relationships between variables and calculate the
probability of each node in the network and the formation of Conditional Probability Tables

52
(CPT). Finally, for network inference was used algorithms as different as Logic Sampling,
Likelihood sampling and EPIS Sampling. The Bayesian Network are shown in Figure 4.
Figure 4- Bayesian network of secondary school students using GeNIe
5. EXPERIMENTAL RESULT ANALYSIS
Using Bayesian Networks obtained, the impact of various factors in field of study can be
examined and offered the appropriate academic major. Finaly, the processing was carried out
using the above-mentioned algorithms, the results are shown in Table 2. Thus we can say that for
test data, the network has the ability to correct diagnosis about 70%. In Bayesian Networks, as
our input data increase, the output of the network show us the results are more likely.
Table 2 -Results of the Bayesian Network using three different algorithms
Algorithm 1 2 3 4 5 6 7 8 9
Likelihood
Sampling
70
% 70
% 70
% 70
% 70
% 72
% 70
% 70
% 78
%
Logic
Sampling
77
% 78
% 77
% 77
% 79
% 78
% 79
% 78
% 78
%
EPIS
Sampling
70
% 73
% 75
% 70
% 73
% 76
% 70
% 70
% 73
%
6. CONCLUSIONS
Bayesian Networks are widely used in detection and prognosis of diseases. The purpose of this
study is to predict the academic major selection for high school students using Bayesian
networks. The effective factors have been used in academic major selection for the first time as an
effective indicator of Bayesian networks According to the survey results of Bayesian Network
outputs with different algorithms using various data, we find that the better outcome for Logic
Sampling algorithm of data.

53
REFERENCES
[1] Paivandi, Saeed. 2012, Februar 2. "Iran: Second Higher Education Cultural Revolution?" Retrieved
From Http://Www.Universityworldnews.Com/Article.Php?Story=20101105221117557
[2] Gottleib, F.M. (2006), "Humanistic Leadership: Emotional Intelligence And Team Learning.
[Dissertation Abstract D.M.]". University Of Phoenix.
[3] Bish, M.R. (2004). "The Impact Of Teacher Pupil Control Ideology On The Development Of Middle
School Self-Efficacy". Unpublished Ed.D Dissertation, University Of Novth Carolina At Charlotte
[4] Singh,S.K.(2007). "Role Of Emotional Intelligence In Organizational Learning : An Empirical
Study". Singapore Management Review.
[5] Kevin B. Korb, Ann E.Nicholson. 2011."Bayesian Artificial Intelligence", Second Edition, Crc Press.
[6] Bayat S., Cuggia M., Et Al., 2008. " Modeling Access To Renal Transplantation Waiting List In A
French Healthcare Network Using A Bayesian Method" Ehealth Beyond The Horizon – Get It There,
605-610.
[7] Yuyang Liu, Wooi Ping Cheah, Byung-Ki Kim, Hyukro Park, “Predict Software Failure-Prone By
Learning Bayesian Network”, International Journal Of Advanced Science And Technology, Pp. 35-
42.
[8] Khlifia Jayech, Mohamed Ali Mahjoub, “Clustering And Bayesian Network For Image Of Faces
Classification”, International Journal Of Advanced Computer Science And Applications, Special
Issue On Image Processing And Analys, May 2011.
[9] Nir Friedman, Michal Linial, Iftach Nachman, Dana Pe'er," Using Bayesian Networks To Analyze
Expression Data", Proceedings Of The Fourth Annual International Conference On Computational
Molecular Biology, Pp. 1-25, 2000.
[10] Pablo Felgaer, Paola Britos, "Prediction In Health Domain Using Bayesian Networks Optimization
Based On Induction Learning Techniques", International Journal Of Modern Physics C, Vol. 17, No.
3, Pp. 447–455 ,2006.
[11] Arezoo Aghaie, Ali Saeedi, "Using Bayesian Networks For Bankruptcy Prediction: Empirical
Evidence From Iranian Companies", International Conference On Information Management And
Engineering, Pp.450-455, 2009.

PREDICTING ACADEMIC MAJOR OF STUDENTS USING BAYESIAN NETWORKS TO THE CASE OF IRAN

More Related Content

Similar to PREDICTING ACADEMIC MAJOR OF STUDENTS USING BAYESIAN NETWORKS TO THE CASE OF IRAN (20)

More from ijcax (20)

Recently uploaded (20)

PREDICTING ACADEMIC MAJOR OF STUDENTS USING BAYESIAN NETWORKS TO THE CASE OF IRAN