ISSN: 2277-9655

[Oloruntoba S.A* et al., 6(12): December, 2017] Impact Factor: 4.116

IC™ Value: 3.00 CODEN: IJESS7

IJESRT
INTERNATIONAL JOURNAL OF ENGINEERING SCIENCES & RESEARCH
TECHNOLOGY
STUDENT ACADEMIC PERFORMANCE PREDICTION USING SUPPORT
VECTOR MACHINE
S.A. Oloruntoba1 ,J.L.Akinode2
Computer Science Department, the Federal Polytechnic,ILARO,Ogun State, Nigeria
Computer Science Department, the Federal Polytechnic,ILARO,Ogun State, Nigeria

DOI: 10.5281/zenodo.1130905

ABSTRACT
This paper investigates the relationship between students' preadmission academic profile and final academic
performance. Data Sample of students in one of the Federal Polytechnic in south West part of Nigeria was used.
The preadmission academic profile used for this study is the 'O' level grades(terminal high school results).The
academic performance is defined using student's Grade Point Average(GPA). This research focused on using
data mining technique to develop a model for predicting student performance based on 'O' level results and their
first 3 semester at each semester. Data preprocessing was done to remove the results of rusticated and expelled
student .Results obtained by comparing SVM with other ML techniques such as KNN,Decision trees, linear
Regression shows that SVM outperforms other ML algorithms. The parameters of the SVM algorithm(kernel)
was also tuned to improve its accuracy and result obtained shows that the RBF kernel with penalty(C=100)
performs best.SVM and RBF gave the highest training accuracy of 94% and 97% predicting accuracy which
outperforms other state of the art ML technique like KNN,decision trees etc

KEYWORDS: Student Perfomance,Prediction,Data Minning,Grade Point Average,SVM.

I. INTRODUCTION
Educational data mining is an emerging research area that aims at developing models for exploring wealth of
data from educational information system. The application of Educational Data Mining(EDM) traversed the
educational sector and has become the new trend in data mining and knowledge discovery in Databases(KDD)
field. it focuses on discovering important patterns and discovering useful knowledge from the academic based
information system[3]. These include course management systems(moodle,blackboard),admissions systems,
registration systems and other systems that are used for managing students at different levels of education from
secondary school to Universities. The process involve in data mining and Knowledge discovery
databases(KDD) is depicted in figure 1.

The main focus of any higher institution is to improve decision making at the managerial level and to impart
education. Sound prediction of students' success in high institutions is one of the basis for improving the quality
of education . Students performance is an important and integral part in higher institutions. This is because the
quality of education in universities is based on its excellent record of academic achievements. Predicting
students performance has become a daunting task due to the large volume of data in educational databases[6].

Educational data mining describes a process used to extract useful information and patterns from a huge
educational database[7]. The extracted information and patterns can be effectively used in predicting students
performance.Hence,this will assist educators at different strata of academic institutions to provide effective
teaching approach and subsequently enhance student's academic performance. Furthermore, educators could
also monitor their students achievements. Similalry,students could improve their learning activities, allowing the
administration to improve the systems performance.

http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology

[588]
 ISSN: 2277-9655
[Oloruntoba S.A* et al., 6(12): December, 2017] Impact Factor: 4.116
IC™ Value: 3.00 CODEN: IJESS7
[5] reiterates that management of academic institutions should focus more on the profile of admitted students,
getting aware of the different types and specific students’ characteristics based on the received data. Similarly,
they should consider if they have all the student's data needed to analyze the students at entry point for decision
making.[5] submit that attributes and prediction methods are the two main factors in predicting students
performances.

There exist several tasks that are used to build a predictive model, these include classification, regression and
categorization. However, most popular task to predict students performance is classification. There are several
data mining algorithms that have been applied to predict students performance. Among the algorithms used are
Decision tree, Artificial Neural Networks, Naive Bayes, K-Nearest Neighbor and Support Vector Machine[6].

This research focused on using data mining technique to develop a model for predicting student performance
based on their 'O' level results and their first 3 semester CGPA (Cumulative Grade Point Average) to predict
their final CGPA.Secondary data was obtained from a Federal Polytechnic in the South West of Nigeria between
2015 and 2016 session . Information’s like student's High School grade(WAEC or O level) and CGPA were
collected from the student's file, to predict the performance at the end of the semester. This paper investigates
the accuracy of SVM for predicting student performance.

The paper is divided into five sections. The purpose and background of the conducted research work is
presented in the Introduction. A quick review of the related research work is provided in Section 2, the
methodology adopted for the research work is described in Section 3, the obtained results and the comparative
analysis are given in Section 4. The paper concludes with a summary of the achievements and discussion of
further work.

Figure 1:Data mining :A KDD process

Source: [22]

http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology

[589]
 ISSN: 2277-9655
[Oloruntoba S.A* et al., 6(12): December, 2017] Impact Factor: 4.116
IC™ Value: 3.00 CODEN: IJESS7

Figure 2. Educational Data Mining Cycle .Source:(Neha Choudary,2016)

II. RELATED WORKS

In this section, we discussed the various approaches for predicting student performance and measurement using
different data mining techniques, documented by various researchers.

[1] was arguably among the foremost authors who classified students by using genetic algorithms to predict
their final grade.

[13] applied Support Vector Machine as their prediction technique in a small datasets.

Using the regression techniques,[2] predicted a student’s marks (pass and fail classes) in their research work

[14] in their paper, predicted a student’s academic success (classified into low, medium, and high risk classes)
using different data mining methods (decision trees and neural network).

[15] used a decision tree model to predict the final grade of group of student who took the C++ course in
Yarmouk University in Jordan.

[25] compares different methods of data mining tools and techniques for classifying students on the basis of
their Moodle usage data and the final score obtained in their corresponding courses. [24] shows the
enhancement of two machine learning model which are used to predict whether a student can answer correct
question in an Intelligent Tutoring system

[16] using the decision tree ,predicted the result of the final exam to help professors determine students who
needed help, in order to improve their performance and obtain pass grade.

[12] stated that Support Vector Machine has a good generalization ability and faster than other methods.

[4] applied three supervised data mining algorithms to assess the data of first year students to predict favorable
outcome in a course and evaluating the performance based on certain factors like convienence,accuracy and
approach of learning.[30] conducted an experimental survey to generate database for students for predicting the
performance. The main focus is to identify the important predictive variables on higher secondary students, to
determine the best algorithm and to predict the grade at higher education

The current research work by [28], compares two tools of data mining applied to data sets of small related to
higher learning institutions and opined that the result will encourage higher education to inculcate data mining
in their business processes. The author successfully predicted the success rate of students' enrolled.

http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology

[590]
 ISSN: 2277-9655
[Oloruntoba S.A* et al., 6(12): December, 2017] Impact Factor: 4.116
IC™ Value: 3.00 CODEN: IJESS7
[18] demonstrated in their research the use of Support Vector Machine to predict risk of failing a course among
student. In their work, Support Vector Machine method acquired the highest prediction accuracy in identifying
students at risk of failing.

[26] developed and tested Support Vector Machine Algorithm and Multiple Linear Regression .The adapted
methodology applied for data set of student enrolled in engineering. The result examined showed that SVM
produce higher accuracy to identify the students having low grades.

[23] studied how the performance of the students evolves during their year of studies. For clustering,
progression of students is used .Thus, students in the same cluster have same progression.

[29] also developed a prediction model that depend on the participation of students through genetic
programming by including learning analytics and educational data mining

[27] developed a methodology to determine the future career of University graduate students. the research aimed
to determine the strategy to improve the performance and scheduling of exams by using different data mining

[9] employed data mining algorithm to predict the student course selection . The research outcome submits that
a students’ grade point average relative to the grades of the courses they are considering for enrolment was the
most important factor in determining future course selections

III. DATA MINING METHODS

Data mining(DM) described a computational method of processing data which is mostly used in many areas
that aim to obtain useful knowledge from the data (Klosgen and Zytkow, 2002). The DM techniques are used to
build a model according to which the unknown data will try to identify the new information. There are several
data mining methods that are used in obtain hidden knowledge from vast amount of data. These include;
Decision theory, Neural Network, Bayesian Classifier-Nearest Neighbor and Support Vector Machine.

Decision Tree
Decision Tree is one of a popular technique for prediction. The technique have been used extensively by most
of researchers because of its simplicity and comprehensibility to uncover small or large data structure and
predict the value. Decision Tree classifiers are used in data mining to produce trees after studying the training
set and will be used to create predictions. Decision tree classifiers are one of the admired and influential tools
for classification. Normally, decision tree classifiers have a tree-like structure which starts from root attributes,
and ends with leaf nodes. It also has several branches consisting of dissimilar attributes, the leaf node on each
branch representing a class or a kind of class distribution. Decision tree algorithms explain the relationship with
attributes, and the comparative significance of attributes. The benefit of decision trees are that they characterize
rules which could simply be understood and interpreted by users, do not need complex data preparation, and do
well for numerical and categorical variables. The core algorithm for constructing decision trees called ID3.
Neural Network
Neural network is arguably one of the popular technique used in educational data mining. The advantage of
neural network is that it has the potential to detect all possible interactions between predictors variables[10] )
.Neural network could also do a complete detection without having any doubt even in complex nonlinear
relationship between dependent and independent variables[17] . Therefore, neural network technique is selected
as one of the best prediction method.

Bayesian Classifier
This is a simple classification method that is based on the theory of probability(the Bayesian
theorem) [17]. It is referred to as naive because it simplifies problems relying on two important
assumptions: it assumes that the prognostic attributes are conditionally independent with familiar
classification, and it supposes that there are no hidden attributes that could affect the process of
prediction. This classifier represents the promising approach to the probabilistic discovery of
knowledge, and it showcase a very efficient algorithm for data classification.

K-Nearest Neighbor
The k-Nearest Neighbor algorithms (k-NN) organize objects based on the neighboring training examples in

http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology

[591]
 ISSN: 2277-9655
[Oloruntoba S.A* et al., 6(12): December, 2017] Impact Factor: 4.116
IC™ Value: 3.00 CODEN: IJESS7
the feature space. K-NN is a kind of instance-based learning, or lazy learning, where the function is only
approximated nearby and the entire calculation is delayed in anticipation of classification. The main problem of
k-NN algorithm is that its accuracy can be strictly ruined by the existence of loud or inappropriate features.
Likewise, its accuracy becomes unfortunate if the feature balance are not reliable with their importance

Support Vector Machine(Svm)

SVMs are described as a set of related supervised learning techniques used for classification and regression [20].
They are member of a family of generalized linear classification. An important property of SVM is , SVM
simultaneously minimize the empirical classification error and maximize the geometric margin. Thus,SVM is
also known as a Maximum Margin Classifiers. SVM is based on the Structural risk Minimization (SRM). SVM
map input vector to a higher dimensional space where a maximal separating hyperplane is constructed. Two
parallel hyperplanes are constructed on each side of the hyperplane that separate the data. The separating
hyperplane is the hyperplane that maximize the distance between the two parallel hyperplanes.

IV. DATA DESCRIPTION

The data for the model were collected from the students' Registration File obtained at the Department of
Computer Science of a Federal Polytechnic in Nigeria for 2015-2016 academic year for graduated students.
After eliminating incomplete data, the sample comprised 89 students who were at the time of researche present
at the practice classes.

The data collected was for 2015 to 2016 session for students who had graduated from the OND(Ordinary
National Diploma). All the predictor and response variables are given in the table 1 below:

As input to the model,12 variables are used, the names and coding are shown in Table 1

Table 1: Student related variables

S/N Variable Name Information Values Data type

1 Eng English Result of the A, B, C, D, E, F Integer
High school terminal
Exam
2 Math Mathematics Result A, B, C, D, E, F Integer
of the High school
terminal Exam
3 Phy Physics Result of the A, B, C, D, E, F Integer
High school terminal
Exam
4 Bio Biology Result of the A, B, C, D, E, F Integer
High school terminal
Exam
5 Agric Agriculture Result of A, B, C, D, E, F Integer
the High school
terminal Exam
6 Eco Economics Result of A, B, C, D, E, F Integer
the High school
terminal Exam
7 CHE Chemistry Result of A, B, C, D, E, F Integer
the High school
terminal Exam
8 A 1st Semester result of Nominal (0-4) Float
a four semester
program
9 B 2nd Semester result Nominal (0-4) Float
of a four semester
program
10 C 3rd Semester result of Nominal (0-4) Float
a four semester

http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology

[592]
 ISSN: 2277-9655
[Oloruntoba S.A* et al., 6(12): December, 2017] Impact Factor: 4.116
IC™ Value: 3.00 CODEN: IJESS7
program
11 D 4th Semester result of Nominal (0-4) Float
a four semester
program

V. METHODOLOGY
Globally, most institution of higher learning adopt grading system to estimate and decide the academic
performance of students.Similarly,we have adopted the same approach for the analysis and measurement of
performance .

Proposed Approach
The initial step is to collect the data set required for the research work. The methodology is applied to a factual
data containing information about the graduated student at the Department of Computer Science at the
Polytechnic. The work flow of this paper is shown in figure 4.

Figure 4:Work flow of Study

Once the data is obtained, it is transformed into required form for mining process, which is called pre-processing
stage. it is an important step used in data mining process and it hinged on transforming the raw data into a
proper format for resolving a particular problem..It has been discovered that the finer the pre-processing is done
of the initial data, the more useful and suitable information is possible to discover.
Immediately, after the data is pre-processed ,we proceed to identify the incomplete ,incorrect and irrelevant data
from our dataset and remove this erroneous and improperly formatted data..This phase is known as data
cleaning phase. This process usually includes eliminating the typing errors or validating and correcting the
valued of entities by cross checking it with accurate data set.

Once the data is complete and consistent in all respects, the next stage is to filter the data according to our
requirement.
[23] highlights the importance of this step
1. Data Fetching: Data fetching combines all the available data that can be used to resolve the data
mining problem, into a set of instances
2. Data Cleanning:At this stage, erroneous and irrelevant data are detected and discarded.
3. Data Filtering: It helps to reduce the large amount of information available to us.
4 Data Transformation: This is the process of deriving new attributes from beforehand available
attributes to assist in a better interpretation of information.

Algorithm Used
The proposed algorithm is the Support vector machine(SVM) which most suitable for small dataset.SVM is the
newest technique for supervised learning. The SVM is used to carry out regression analysis on the ready data
set.

http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology

[593]
 ISSN: 2277-9655
[Oloruntoba S.A* et al., 6(12): December, 2017] Impact Factor: 4.116
IC™ Value: 3.00 CODEN: IJESS7

Figure5 Support Vector Machine [25]

SVM Algorithm:
i. Designate an optimal hperplane to maximize the margin
ii. Widen the above definition for non -linear separable problems
iii. Map the data to high dimensional space where it is simple to classify with linear decision and
reformulate problem so that data is mapped completely to this space.

VI. IMPLEMENTATION AND RESULT

Preprocessing of the data
Missing values of an attribute was cleaned by filling the value with average of fields on attribute. Meanwhile,
the missing values that found at class involve that all field corresponding with that field of class was deleted. To
simplify prediction process, GPA is used as the response variable.

Figure 6: The Box plot of the sample data

http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology

[594]
 ISSN: 2277-9655
[Oloruntoba S.A* et al., 6(12): December, 2017] Impact Factor: 4.116
IC™ Value: 3.00 CODEN: IJESS7

Figure 7: Histogram of each attribute

The histogram shows that there is strong correlation between the attributes

Figure 8: Scatter matrix

The scatter matrix also shows that there is strong correlation between the attribute

Experimental Result
The software chosen for the analysis was python Sklearn. Sklearn provides state of the art classification and
regression algorithms such as linear Regression, K-nearest neighbor, Decision trees, Support Vector Regression.
The dataset was divided into 70% training and 30% testing following[21].

http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology

[595]
 ISSN: 2277-9655
[Oloruntoba S.A* et al., 6(12): December, 2017] Impact Factor: 4.116
IC™ Value: 3.00 CODEN: IJESS7
Four regression algorithms were used to analyze the dataset. These include ; Linear Regression (LR), Lasso
Regression (LASSO) and ElasticNet (EN), Classification and Regression Trees (CART), Support Vector
Regression (SVR) and k-Nearest Neighbors (KNN).
All the algorithms use their default tuning parameters. Comparing the algorithms and displaying the Mean
Squared Error (MSE) and standard deviation for each algorithm was shown below

S/N Algorithm MSE Standard Deviation

1 Linear Regression -0.042915 0.021999
5 Decision tree -0.083867 0.051384
6 SVR -0.170131 0.059279
2 LASSO -0.157838 0.059085

3 Elastic Net -0.157838 0.059085

4 KNN -0.119799 0.038263
Table2: Comparative analysis of algorithms using MSE and Standard deviation

From the table above, SVR has the lowest MSE which shows the wrong predictions all the algorithms are
performing (0 is perfect).
The parameters of the Support Vector Regression (SVR) are now tuned to improve the classification accuracy
such as the penalty parameters (C) and kernel function. The percentage accuracy of the different kernels with C
= 10, 100, 1000
C= 10
S/N Kernel Training Testing
1 Linear 77 79
2 Poly 94 65
3 RBF 94 97
Table 3: Accuracy of different SVM Kernel when C=10
C= 100
S/N Kernel Training Testing
1 Linear 78 75
2 Poly 94 51
3 RBF 94 98
Table 4: Accuracy of different SVM Kernel when C=100
C= 10000
S/N Kernel Training Testing
1 Linear 45 49
2 Poly 94 51
3 RBF 94 96
Table 5: Accuracy of different SVM Kernel when C=1000

VI. CONCLUSION AND FUTURE WORK

In this paper, Support Vector Machine data mining algorithm was applied to predict student success at the end
of their study based on their 'O' level result and CGPA obtained at each semester Predicting student
performance can be useful to the managements in many contexts. For identifying excellent students for
scholarship programs, admissions, and to help level advisers to quickly identify students who are unlikely to
graduate. From the results it is proven that Support vector regression algorithm remains the state of the art
algorithm for predicting student performance. SVM gives 98% prediction for 89 instances which is relatively
higher than other classifier and the MSE error rate is very low. For future work, an efficient method for tuning
the performance of the algorithm such as particle swarm optimization could even provide better results

http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology

[596]
 ISSN: 2277-9655
[Oloruntoba S.A* et al., 6(12): December, 2017] Impact Factor: 4.116
IC™ Value: 3.00 CODEN: IJESS7

VII. REFERENCES
1. B. Minaei-Bidgoli ; D.A. Kashy ; G. Kortemeyer ; W.F. Punch,2003,Predicting student
performance: an application of data mining methods with an educational Web-based
system Frontiers in Education, 2003. FIE 2003 33rd Annual
2. S.B. Kotsiantis ; P.E. Pintelas, 2005,Predicting students marks in Hellenic Open University
Advanced Learning Technologies, 2005. ICALT 2005. Fifth IEEE International Conference on
3. Amjad Abu Saa,2016,Educational Data Mining & Students’ Performance Prediction Information
Technology Department Ajman University (IJACSA) International Journal of Advanced Computer
Science and Applications, Vol. 7, No. 5, 2016 212 | P a g e www.ijacsa.thesai.org
4. Edin Osmanbegović &, Mirza Suljić ,2012,Data Mining Approach For Predicting Student
Performance Economic Review – Journal Of Economics And Business, Vol. X, Issue 1, May 2012
5. Dorina Kabakchieva ,2013, Predicting Student Performance By Using Data Mining Methods For
Classification Cybernetics And Information Technologies
6. Amirah Mohamed Shahiri , Wahidah Husain , Nur’aini Abdul Rashid ,2015, A Review on Predicting
Student’s Performance using Data Mining Techniques, Procedia Computer Science 72 ( 2015 ) 414 –
422 Available online at www.sciencedirect.com
7. Angeline D. M. D., 2013,Association rule generation for student performance analysis using apriori
algorithm, The SIJ Transactions on Computer Science Engineering & its Applications (CSEA) 1 (1)
(2013) p12–16.
8. Ankita Katare and Shubha Dubey2017, A Study of various Techniques for Predicting student
Performance under Educational Data Mining International Journal of Electrical, Electronics ISSN
No. (Online): 2277-2626 and Computer Engineering 6(1): 24-28(2017)
9. Ognjanovic ,D. Gavesic and,S. Dawson , 2016,Using insttutional data to predict student course
selections in higher education ,The internet and Higher Education,vol 29 ,pp 49-62
10. G. Gray, C. McGuinness, P. Owende, An application of classification models to predict learner
progression in tertiary education, in: Advance Computing Conference (IACC), 2014 IEEE
International, IEEE, 2014, pp. 549–554.
11. I. Hidayah, A. E. Permanasari, N. Ratwastuti, Student classification for academic performace
12. S. Sembiring, M. Zarlis, D. Hartama, S. Ramliana, E. Wani, 2011,Prediction of student academic
performance by an application of data mining techniques, in: International Conference on
Management and Artificial Intelligence IPEDR, Vol. 6, 2011, pp. 110–114
13. W. Hamal ainen, ¨ M. Vinni, 2006, Comparison of machine learning methods for intelligent tutoring
systems, in: Intelligent Tutoring Systems, Springer, 2006, pp. 525–534.
14. Superby, J. Vandamme, J., Meskens, N. (2006). Determination of factors influencing the
achievement of the first-year university students using data mining methods. Proceedings of the
Workshop on Educational Data Mining at the 8th International Conference on Intelligent Tutoring
Systems (ITS 2006). Jhongli, Taiwan, pp37-44.
15. Al-Radaideh, Q., Al-Shawakfa, E. & AlNajjar, M. (2006), Mining Student Data Using Decision
Trees, International Arab Conference on Information Technology (ACIT'2006), Yarmouk
University, Available on: http://titania.addu.edu.ph/researches/DE
CISION%20SUPPORT/Mining%20Student %20Data%20Using%20Decision%20Trees .pdf [pristup
10.januar 2012.]
16. Kumar S. A. & Vijayalakshmi M. N. (2011), Efficiency of Decision Trees in Predicting Student's
Academic Performance, First International Conference on Computer Science, Engineering and
Applications, CS and IT 02, Dubai, pp. 335-343.
17. P. M. Arsad, N. Buniyamin, J.-l. A. Manan,2013, A neural network students’ performance prediction
model (nnsppm), in: Smart Instrumentation, Measurement and Applications (ICSIMA), 2013 IEEE
International Conference on, IEEE, 2013, pp. 1–5.
18. G. Gray, C. McGuinness, P. Owende, 2014,An application of classification models to predict learner
progression in tertiary education, in: Advance Computing Conference (IACC), 2014 IEEE
International, IEEE, 2014, pp. 549–554
19. Witten, I.H. & Frank E. (2000), Data Mining – Practical Machine Learning Tools and Techniques,
Second edition, Morgan Kaufmann, San Francisco.
20. V. Vapnik.1995, The Nature of Statistical Learning Theory. NY: Springer-Verlag.

http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology

[597]
 ISSN: 2277-9655
[Oloruntoba S.A* et al., 6(12): December, 2017] Impact Factor: 4.116
IC™ Value: 3.00 CODEN: IJESS7
21. Ritesh Agr a wal ,2 0 1 2 ,Dividing Data Into Training And Testing In R
https://ragrawal.wordpress.com/2012/01/14/dividing-data-into-training-and-testing-dataset-in-r/
retrieved on 19th Dec.2017
22. Jiawei Han and Micheline Kamber,2006 ,Data Mining: Concepts and Techniques, 2nd ed. The
Morgan Kaufmann Series in Data Management Systems, Jim Gray, Series Editor
Morgan Kaufmann Publishers, March 2006. ISBN 1-55860-901-6
23. R .Asif. A., and M.K Pathan, 2014, Predicting student academic performance at degree level: A
case study,International Journal of Intelligient Systems and Applications vol 7 no1, pp. 49-61.
24. Mavrikis M,,2008,Data-driven modelling of student's interactions in an ILE In EDM,pp.87-96
25. Romero, C. Ventura ,P.G Espejo C,Hervas,Data,2008, Minning Algorithms to Classify Students,In
EDM, pp.8-17
26. John M. Mativo& Shaobo Huang,2014, Prediction of students' academic performance: Adapt a
methodology of predictive modeling for a small sample size", , vol. 00, no. , pp. 1-3,
doi:10.1109/FIE.2014.7044287
27. Renza Campagni,Donatella Merlini, Renzo Sprugnoli,Maria CeciliaVerri ,2015,Data mining models
for student careers,Expert Systems with Applications, Volume 42, Issue 13, 1 August, Pages 5508-
5521
28. Srecko Natek a&, Moti Zwilling,2014 Student data mining solution–knowledge management system
related to higher education institutions, Expert Systems with Applications 41 (2014) 6400–6407
29. Xing Wanli , Guo Rui, Petakovic Eva , Goggins Sean,2015 Participation-based student final
performance prediction model through interpretable Genetic Programming: Integrating learning
analytics, educational data mining and theory, Computers in Human Behavior 47 (2015) 168–181
30. V.Ramesh, P.Parkavi & K.Ramar,2013, Predicting Student Performance: A Statistical and Data
Mining ApproachInternational Journal of Computer Applications (0975 – 8887) Volume 63– No.8

CITE AN ARTICLE
Oloruntoba, S. A., & Akinode, J. L. (n.d.). STUDENT ACADEMIC PERFORMANCE
PREDICTION USING SUPPORT VECTOR MACHINE. INTERNATIONAL JOURNAL OF
ENGINEERING SCIENCES & RESEARCH TECHNOLOGY, 6(12), 588-598.

http: // www.ijesrt.com© International Journal of Engineering Sciences & Research Technology

[598]