Estimated Costs for Single Tuition Fee (STF) using Naïve Bayes

Method

Juhriyansyah Dalle

, Dwi Hastuti

, Taufik Rahman

, A. Akrim

, Sri Erliani

Taufik Hidayat

, Siska Devina

, Agustina Lestari

, B. Baharuddin

, Hesti Fibriasari

Akhmad Murjani

, Erika Lismayani

, Ahmad Yusuf

10 m

and Candra Kusuma Negara

Department of Information Technology, Universitas Lambung Mangkurat, Jl. H. Hasan Basry, Banjarmasin, Indonesia

Department of Electrical Engineering, Universtias PGRI Adi Buana, Jl. Ngagel Dadi III No.3B/37, Surabaya, Indonesia

Department of Higher Education Management, Universitas Muhammadiyah Sumatera Utara, Jl. Kapten Muchtar Basri

No.3, Medan, Indonesia

Department of Public Health, Universitas Cahaya Bangsa, Jl. A. Yani Km. 17, Banjarmasin, Indonesia

Department of Management, Universitas Cahaya Bangsa, Jl. A. Yani Km. 17, Banjarmasin, Indonesia

Department of Nursing Sciences, Universitas Cahaya Bangsa, Jl. A. Yani Km. 17, Banjarmasin, Indonesia

Department of Electrical Engineering Education, Universtias Negeri Medan, Jl. Wiliem Iskandar Pasar V, Medan,

Indonesia

Department of French Education, Universtias Negeri Medan, Jl. Wiliem Iskandar Pasar V, Medan, Indonesia

Department of Law, Universitas Cahaya Bangsa, PSDKU, Jl. Pemuda, Kuala Kapuas, Indonesia

Department of Information Technology, Universitas Cahaya Bangsa, , Jl. A. Yani Km. 17, Banjarmasin, Indonesia

baharuddin.bah, akhmad.murjani9, erikalismayani30}@gmail.com, akrim@umsu.ac.id, hesti@unimed.ac.id,

ahmadyusuf23.ay@gmail.com, candra14780@yahoo.com

Keywords: Decision Support System, STF, Naïve Bayes.

Abstract: The difference in the amount of single tuition fee (STF) paid by students with middle and upper economic

backgrounds causes an injustice gap. This is partly due to the instability of the system used, especially in

terms of STF determination methods. Other additional shortcomings include the criteria entered the system

that is still not enough to be considered in determining STF. Therefore, this study aims to build a web based

STF payment system using the Naïve Bayes, probability, and statistical methods for students to determine the

cost of an institution's tuition fee easily. System testing is carried out by comparing the output and verification

results. This showed that the estimated determination of the cost of STF payments is suitable, with 83.3%.

https://orcid.org/0000-0000-5700-2766

https://orcid.org/0000-0001-8716-7062

https://orcid.org/0000-0002-5653-2298

https://orcid.org/0000-0002-3984-9242

https://orcid.org/0000-0002-0662-3175

https://orcid.org/0000-0001-9659-4890

https://orcid.org/0000-0002-0075-6911

https://orcid.org/0000-0003-1180-9625

https://orcid.org/0000-0001-8334-3030

https://orcid.org/0000-0002-8009-4603

https://orcid.org/0000-0002-1162-2660

https://orcid.org/0000-0002-7963-4852

https://orcid.org/0000-0003-2383-9944

https://orcid.org/0000-0002-9831-7465

280

Dalle, J., Hastuti, D., Rahman, T., Akrim, A., Erliani, S., Hidayat, T., Devina, S., Lestari, A., Baharuddin, B., Fibriasari, H., Murjani, A., Lismayani, E., Yusuf, A. and Negara, C.

Estimated Costs for Single Tuition Fee (STF) using Naïve Bayes Method.

DOI: 10.5220/0010476702800289

In Proceedings of the 23rd International Conference on Enterprise Information Systems (ICEIS 2021) - Volume 1, pages 280-289

ISBN: 978-989-758-509-8

1 INTRODUCTION

The tuition fee of tertiary institutions is one of the

main aspects considered by prospective students

when choosing to further their studies. According to

Gyamfi, Gyamfi, & Qi (2016), tuition fees are

compared with the product, place, process, student

quality, and graduate evidence. Ozekicioglu (2017)

stated that living costs are also considered in the

financial calculations of prospective students. In

contrast, Bates & Kaye (2014) reported that changes

in the amount of payment directly affect the number

of applicants at tertiary institutions. Generally, when

tuition fees are increased, the number of applicants

tends to decrease, and vice versa (Sulistiyo & Soegoto,

2018). However, according to a study carried out by

Burgess, Senior, & Moores (2018), the possibility of

an increase in tuition fees can be minimized by

improving the quality of service and providing a

general view of universities. Lassila (2011) stated that

providing financial assistance also enables students

from low-income families to further their studies into

tertiary institutions.

Therefore, a policy regarding the proportional

tuition fee, which does not disrupt the student's

family's financial flow, is needed (Surtiati, Siregar, &

Andati, 2017). This led to the evolution of the STF

policy, which was applied in accordance with the

Minister of Education and Culture Regulation

Number 55 of 2013. This was further amended to

Minister of Education and Culture Regulation

Number 73 of 2014, which was based on an

evaluation of the implementation of single tuition fees

for the 2013-2014 academic years at state universities.

Additional changes occurred in 2015 and 2016

when the law was amended to No. 22 of 2015 and No.

39 of 2016, which stated that the costs borne by

students need to be adjusted to the economic

capabilities of their parents, or guidance. Furthermore,

legal certainty in determining the costs borne by

students need to be arranged regarding the cost of

single tuition at state universities. In 2017 the law was

amended by the Minister of Education and Culture

Regulation to No. 39 of 2017 to regulate the costs

borne by students according to their parents'

economic capabilities or other parties that financed

their education. This aimed at arranging a single

tuition fee on state universities within the Ministry of

Research, Technology, and Higher Education

(Kemenristek-Dikti, 2017).

Dunga & Mncayi (2016) stated that improving the

payment system is an absolute thing to do in order to

assist students. However, this is not always directly

proportional to enhancing the quality of higher

education, made for the principles of accountability

(Fortunata & Toni, 2020). This law was also amended

to enable tertiary institutions to pay attention to

aspects of parents' economic ability on an ongoing

basis (Sumarno, Gimin, & Nas, 2017). This is also to

prevent any form of protest likely to occur due to the

unilateral determination by the stakeholders (Kajawo,

2019).

Implementing the STF system is to ease the

financial burden of education funding on students

(Fauzi, 2017). STF is the amount of fees that need to

be paid by students each semester. It is divided into

several groups, and each state university has 5

divisions, with some having more such as Indonesia

Educational University.

Universitas Lambung Mangkurat Banjarmasin

adopted the STF System comprising of 5 groups in

their 2013 academic year. With this system, tuition

fees are paid every semester without building fees

(Universitas Lambung Mangkurat, 2017). During this

time, the university had a system to determine the

class of STF. However, the required criteria were still

lacking and did not use methods/algorithms to make

calculations, therefore, the obtained results were

inadequate in decision making. Besides that, with the

existing system, prospective students are unable to

determine the estimated cost of payment to the

applied tertiary institution, which is very important

(Galvin, Nieuwnhuis, Phillips, Thain, & Kokkori,

2015).

Furthermore, Wardi, Abror, & Trinanda (2018),

stated that the current STF system was inadequate and

had a significant effect on students' desire to move to

tertiary institutions. Therefore, for this reason, a web-

based system for determining the cost of payment for

STF was developed. This system uses a classification

method known as the Naive Bayes, which comprises

probability and statistical methods to predict future

opportunities based on past experiences. This is also

known as the Bayes Theorem.

The use of the Naïve Bayes method evolved due

to the predetermined use of data in various fields all

over the world (Siledar & Chaudhary, 2017). Zhang

and Gao (2011) used the Naïve Byes method in text

classification based on the probability of students'

ability in class. Furthermore, Safri, Arifudin, &

Muslim (2018) used it as a fundamental method to

determine those that have the right to obtain a health

insurance card provided by the state.

The Naïve Bayes method is also currently used in

the field of economics and education. For example,

Lagman et al. (2019) used this method to predict the

speed and number of students graduating from

college. Mirza (2019), also used this method to

Estimated Costs for Single Tuition Fee (STF) using Naïve Bayes Method

281

determine the right marketing and promotion

strategies to attract new students. In economics,

Kricehene proved that this method can be used to

determine the probability of a bank's financial risk

(2017). Therefore, the Naïve Bayes method can be

used to determine the amount of single tuition for

students in tertiary institutions.

With this system, prospective students can

determine the estimated amount of Fee to be paid

before the rector for proper verification.

2 METHODOLOGY

This research was carried out in five stages, namely

planning, analysis, design, implementation, and

system.

2.1 Planning

This is the earliest stage carried out to determine the

purpose of making the system. This stage involves

data collection, study libraries, interviews, and

observations.

A literature study is carried out as a reference,

source, and theories in research. The references in this

study are journals, e-books, books, and the internet

(electronic media).

Interviews stage is carried out with the directorate

concerned on the STF that is the finance department.

The topics discussed were, related to the definition of

STF, how it is determined, and the various obstacles

at Universitas Lambung Mangkurat.

After the interview stage, this process is carried

out to determine the registration process and fees for

prospective new students. Observation results were

used to make the system, which is later used to form

the data decision support system for STF.

2.2 Analysis

The analysis phase is carried out to analyze,

understand, and provide a detailed document of the

problems, solutions, and system requirements. It

consists of a system, problem, and data analysis.

Systems analysis is an inseparable part of the

study of business (Fotache, Olaru, & Iacoban, 2015),

and software products. It is useful for an in-depth

study of the needs of a product for suitable usage

(Logunova et al., 2018). Analysis of the existing

system enables Universitas Lambung Mangkurat to

determine the class of STF. However, the criteria

entered into the system are still lacking, and

methods/algorithms are not used in making

calculations, therefore, the results are still inadequate

in making decisions.

The problem is analyzed to identify and solve the

difficulties encountered (Annamalai, Kamaruddin, &

Azid, 2013). Vizioli & Kaminski (2017) stated that

understanding the problem also means providing in-

depth descriptions. Innovative solutions are provided,

assuming problem analysis is appropriate (Kim, Choi,

Chang-Soo, & Park, 2018). Currently, prospective

new students are unable to determine the amount of

STF payment fees based on the criteria of the applied

tertiary institution. Therefore, for this reason, a web-

based STF fee payment estimation system was

developed to determine the estimated STF fee for

prospective new students. This system uses a

classification method known as the Naive Bayes,

which comprises the probability and statistical

methods to predict future opportunities based on past

experiences.

This stage is used to analyze the data needed in

this research case study, such as parents' income,

father's income, mother's income, PLN (electric

power), number of families, number of dependents,

and Indonesia Smart Card. These parameters are

shown in Appendix 1.

A total of 400 data from the university and 40

from the manual calculation examples were used in

this research (See appendix 2).

Data testing aims to determine the level of the

presentation obtained by the Naive Bayes method in

knowing how accurate the method is in solving STF

problems. It was obtained from 30 data.

2.3 Design

At this stage, the tools used for system requirements,

namely a computer/laptop and the software part is

XAMPP, notepad, and web browser were determined.

In addition, a flowchart or UML system was used to

make the system.

This study's research material includes the results

of surveys and observations carried out at Universitas

Lambung Mangkurat. The tool used to implement the

decision support system in this study is a computer,

and the software part is XAMPP, notepad, and web

browser.

The section comprises of planning or drawing a

sketch that functions for the concept of making the

system. The design can be carried out by developing

a flowchart or UML system.

ICEIS 2021 - 23rd International Conference on Enterprise Information Systems

282

Figure 1: Activity Diagram of Prospective Students.

Figure 2: Admin Activity Diagram.

2.4 Implementation

This phase translates the design results in the previous

stage into a system with a programming language. At

this stage, a system is designed to estimate the cost of

determining STF web-based payments for

prospective new students.

2.5 System

At this stage, a system is formed, and testing is carried

out to ensure it is made according to the study's

objectives and needs. Furthermore, at this stage, a

number of tests, including the Verifier Test, were

carried out to determine the STF class accuracy.

Black box Testing is a functional test used to test

software without knowing the internal structure of the

code or program. Usability Testing is one way to

determine whether users can easily use an

application. It is also used to determine a program’s

efficiency and effectiveness and help users achieve

their goals.

3 RESULTS AND DISCUSSIONS

In the following example, it is assumed that the data

is taken from a user and entered into the system. Table

1 shows the training and prospective student data

selected based on the classification criteria of the STF

class.

Approximately 400 training data were used to

perform Naive Bayes calculations, as shown in table 1.

The Naive Bayes method, used the following

stages of calculation, namely: (1) counting the

number of classes / labels, (2) count the same number

of cases with the same class, (3) multiplying all

results of group variables 1, 2, 3, 4, and 5, (4)

comparing the high probability value. It found that the

highest probability value is 0.005, which means it is

included in group 3 (See appendix 3).

Table 1: Case Studies.

3.1 System Process

STF classification results were obtained from data

collection and manual processing using the Naive

Bayes method. This was followed by moving the

database into the system in order to classify user

entries.

The following is the process of calculating the

system, starting with entering data according to

criteria for proper classification in order to provide

results in the form of STF classes.

Admin System Database

Me nu

Sele ction

Vie w Data

Take Data

Display

SNPTN/SBMPTN

Aspects Note

1 Parents' incomes 2.000.000 – 2.500.000

2 Father's income 2.500.000

3 Mother's income 0

4 Electrical power 450 kwh

5 Number of families 5

6 Number of dependents 4

7 Indonesia smart card (kip) None

Estimated group ?

Estimated Costs for Single Tuition Fee (STF) using Naïve Bayes Method

283

Figure 3: Main Page.

3.2 Main Page

This page consists of 3 menus in the header section,

namely STF, New STF, and Login. The STF menu

functions by checking the results of the proof that has

been processed, while the new STF menu serves as

the input of STF data for prospective students of

ULM to determine their STF group

recommendations. The last is the login menu, which

enables admins to log in to the back end of the website

to manage STF data of registered prospective ULM

students.

Figure 4: STF New Filling Page.

3.3 STF New Page

On the new STF page, prospective students are asked

to fill in their data for the system to determine their

estimated class.

First, the user fills in their personal data along

with information on their "Parents' Income, Father's

Income, Mother's Income, Electricity (PLN), Number

of Families, Number of Dependents, and Smart

Indonesia Cards."

After filling out the user selects the process button

and classifies the data using the Naive Bayes

algorithm as follows:

1. Count the number of classes/Labels.

2. Count the same number of cases with the same

class.

3. Multiply all the results of Group 1, 2, 3, 4, and 5

variables.

4. Then compare and determine the highest

probability value.

5. Store in a database and the system output in the

form of STF results.

Figure 3 is a display of the results of the output

system, which has carried out the Naive Bayes

classification process.

Figure 5: Preview Results from the System.

3.4 Verification Test

Verification testing is carried out by comparing the

results issued by the system with the verifier. The

formulas used for testing system results are as

follows:

Total Correc

Total Data

 100%  accurac

results

)

The test results used to estimate the cost of STF

payments using the Naive Bayes method at

Universitas Lambung Mangkurat is calculated as

follows:

Accurac

results 

 100%  83,3%

)

Therefore, the percentage level of conformity on

the system’s outputs with the verified results is

83.3%.

The explanation shows that several factors

influence the calculation to determine the amount of

single tuition. The first is the family’s economic

background, which is the basis for determining the

amount of a single tuition payment cluster and

maintaining inequality to ensure the interest of going

to college is maintained (Tang, Tang, & Tang, 2004).

Previous studies have shown that cluster efficiency

helps maintain the quantity and perception of students

that register at a tertiary institution. Burer & Fethke

(2016) stated that the distribution of payments in the

cluster encourages students to pay for tuition.

ICEIS 2021 - 23rd International Conference on Enterprise Information Systems

284

Indirectly, the quantity of students’ impacts on

improving the physical quality of the campus, leads

to good maintenance and makes the instructors

serious (Wanjala & Ali, 2017). Lin (2016) reported

that the suitability of the amount of payment with

students' economic ability directly impacts their

learning abilities and their desire to achieve academic

success.

The next key factor is the use of the Naïve Bayes

method, a key factor due to its ability to classify data

in the appropriate cluster (Wibawa et al., 2019).

Kaviani & Dhotre (2017) stated that the use of

algorithms in this method overrides subjective

assumptions and makes calculations more accurate.

In addition, it has proven to be able to prevent

corporations, institutions, and industries from

financial risk due to incorrect calculation of funds

received (Jang, Lee, Lee, & Han, 2015). Not only in

the field of finance, the use of Naïve Bayes is also

able to accurately predict the number of diligent

students according to their performance (Shaziya,

Zaheer, & G.Kavitha, 2015).

The benefits provided from the use of naïve Bayes

have proven to be able to provide adequate assistance

for the main educational institutions in calculating

costs (Makhtar, Nawang, & Shamsuddin, 2017). In

addition, the implementation of the Electronic

information system facilitates students and operators

in determining STF. Due to the current digital era, the

use of information technology is a necessity in all

fields (Dalle et al., 2015). Therefore, further research

is needed on a broader study of the use of Naïve

Bayes in educational institutions due to the limited

scope of this study.

4 CONCLUSION

The following conclusions are drawn from the

research carried out by observing and testing the

various stages of the decision support systems:

1. Naive Bayes method can be applied to estimate

and determine the cost of STF payments at

Universitas Lambung Mangkurat.

2. The classification process is carried out using the

Naïve Bayes method, which works by classifying

probabilities and statistics based on previous data.

3. The results issued or recommended by the system

are suitable and in accordance with 83.3%.

ACKNOWLEDGEMENTS

We are thankful to employee of main office of

Universitas Lambung Mangkurat who facilitate us in

collecting data especially for single tuition fee. Also,

we thankful to Rector Universitas Cahaya Bangsa,

Rector Universitas Negeri Medan, Dean of Faculty of

Engineering ULM, and Head of Department of

Information Technology who gave us more space for

discussion of the paper.

REFERENCES

Annamalai, N., Kamaruddin, S., & Azid, I. A., 2013.

Importance of problem statement in solving industry

problems. Applied Mechanics and Materials, 421, 857-

863. doi: https://doi.org/10.4028/www.scientific.net/

AMM.421.857

Bates, E. A., & Kaye, L. K., 2014. Exploring the Impact of

the Increased Tuition Fees on Academic Staffs’

Experiences in Post-92 Universities: A Small-Scale

Qualitative Study. Education Sciences, 4, 229-246.

doi:10.3390/educsci4040229

Burer, S., & Fethke, G., 2016. Nearly-efficient tuitions and

subsidies in American public higher education.

Economics of Education Review, 5, 182-197. doi:

http://dx.doi.org/10.1016/j.econedurev.2016.09.003

Burgess, A., Senior, C., & Moores, E., 2018. A 10-year case

study on the changing determinants of university

student satisfaction in the UK. PLOS ONE, 13(2), 1-9.

doi: https://doi.org/10.1371/journal.pone.0192976

Dalle, J., Mutalib, A. A., Saad, A. L., Ayub, M. N., Wahab,

A. W., & Nasralla, A. M., 2015. Usability

considerations make digital interactive. Journal of

Technology (Sciences & Engineering), 77(29), 63–68.

doi:10.11113/jt. v77.6837

Dunga, S. H., & Mncayi, P., 2016. Determinants of the

perceptions of free higher education among students at

a south african university. International Journal of

Economics Aad Finance Studies, 8(1), 161-176.

Fauzi, D. R., 2017. Retrieved from Kompasiana:

https://www.kompasiana.com/dadanrizwan/58b31b18

b27e61bc0a4f7f35/dinamika-penerapan-sistem-uang-

kuliah-tunggal-ukt-di-ptn

Fortunata, R., & Toni, N., 2020. The influence of service

quality, tuition fee and promotion on student decisions

to continue study in postgraduate of prima indonesia

university. Budapest International Research and Critics

Institute-Journal, 3(2), 829-837. doi: https://doi.org/10.

33258/birci.v3i2.910

Fotache, M., Olaru, G., & Iacoban, V., 2015. Information

systems analysts and designers.Academic/research

supply vs market demand. Procedia Economics and

Finance, 20, 232-242. doi:doi: 10.1016/S2212-

5671(15)00070-2

Galvin, J., Nieuwnhuis, M., Phillips, B., Thain, L., &

Kokkori, C., 2015. Students’ perceptions of tuition fees:

Estimated Costs for Single Tuition Fee (STF) using Naïve Bayes Method

285

an interpretative phenomenological analysis. British

Journal of Education, Society & Behavioural Science,

10(2), 1-15. doi: 10.9734/BJESBS/2015/18981

Gyamfi, E., Gyamfi, I., & Qi, D. K., 2016. The Influence of

Tuition Fees on the Choice of University among Senior

High School Graduates in Kumasi, Ghana.

International Journal of Business and Management,

11(12), 222-229. doi: 10.5539/ijbm. v11n12p222

Jang, W., Lee, J. K., Lee, J., & Han, S. H., 2015. Naïve

bayesian classifier for selecting good/bad projects

during the early stage of international construction

bidding decisions. Mathematical Problems in

Engineering, 1-12. doi: https://doi.org/10.1155/2015/

830781

Kajawo, S. C., 2019. Analyzing members of public "s

reaction to the 2015/2016 fees hike on higher education

in public universities in malawi. Mediterranean Journal

of Basic and Applied Sciences, 3(2), 218-229.

Kaviani, P., & Dhotre, S., 2017. Short survey on naive

bayes algorithm. International Journal of Advance

Engineering and Research Development, 4(11), 607-

611.

Kemenristek-Dikti., 2017. Single tuition fees and single

tuition fees at the State Universities in the Ministry of

Research, Technology and Higher Education. State

News, 779, 1-7.

Kim, J. Y., Choi, D. S., C.-S. S., & Park, J. Y., 2018. The role

of problem solving ability on innovative behavior and

opportunity recognition in university students. Journal of

Open Innovation: Technology, Market, and Complexity,

4(4), 1-13. doi:10.1186/s40852-018-0085-4

Krichene, A., 2017. Using a naive Bayesian classifier

methodology for loan risk assessment:Evidence from a

Tunisian commercial bank. Journal of Economics,

Finance and Administrative Science, 22(42), 3-24. doi:

https://doi.org/10.1108/JEFAS-02-2017-

0039Permanent link to this document:

https://doi.org/10.1108/JEFAS-02-2017-0039

Lagman, A. C., Calleja, J. Q., Fernando, C. G., Gonzales, J.

G., Legasp, J. B., Ortega, J. H., . . . Santos, R. C., 2019.

Embedding Naïve Bayes Algorithm Data Model in

Predicting Student Graduation. ICTCE, 51-56. doi:

https://doi.org/10.1145/3369555.3369570

Universitas Lambung Mangkurat., 2017. Info on Tuition

Fees. Retrieved from Universitas Lambung Mangkurat:

http://biayakuliah.net/biaya-kuliah-universitas-

lambung-mangkurat-unlam-banjarmasin/

Lassila, N. E., 2011. Effects of tuition price, grant aid, and

institutional revenue on low-income student enrollment.

Journal of Student Financial Aid, 41(3), 24-41.

Lin, C.-H., 2016. Examining the effects of financial aid on

student persistence in taiwanese higher education.

International Education Studies, 9(8), 36-50. doi:

10.5539/ies.v9n8p36

Logunova, O., Yu V., K., Torchinsky, V., Sibileva, N.,

Arefyeva, D. Y., & Popov, S., 2018. Usage of system

analysis methods in the software products engineering.

ARPN Journal of Engineering and Applied Sciences,

13(9), 3294-3298.

Makhtar, M., Nawang, H., & Shamsuddin, S. N., 2017.

Analysis on students performance using naïve bayes

classifier. Journal of Theoretical and Applied

Information Technology, 95(16), 3993-4000.

Mirza, A. H., 2019. Application of naive bayes classifier

algorithm in determining new student admission

promotion strategies. Journal of Information System

and Information, 1(1), 14-28. Retrieved from

ttp://journal-isi.org/index.php/isi

Ozekicioglu, S., 2017. Students’ Economic Backgrounds in

Meeting Living Costs of Higher Education: A

Regression Analysis. International Journal of

Economics and Financial Issues, 7(3), 161-165.

Safri, Y. F., Arifudin, R., & Muslim, M. A., 2018. K-

Nearest neighbor and naive bayes classifier algorithm

in determining the classification of healthy card

indonesia giving to the poor. Scientific Journal of

Informatics, 5(1), 9-18. Retrieved from

https://journal.unnes.ac.id/nju/index.php/sji

Shaziya, H., Zaheer, R., & G.Kavitha., 2015. Prediction of

students performance in semester exams using a naïve

bayes classifier. International Journal of Innovative

Research in Science, Engineering and Technology, 4(10),

9823-9829. doi: 10.15680/IJIRSET.2015.0410072

Siledar, S. B., & Chaudhary, S. R., 2017. Comparative

analysis of naive bays classifier and decision tree c4.5

on credit payment data set. International Journal of

Research in Engineering and Technology, 6(4), 43-44.

Sulistiyo, D., & Soegoto, H. S., 2018. The impact of tuition

fees change on the university new student aplicants.

International Journal of Economics, Business and

Management Research, 2(6), 81-87.

Sumarno, Gimin, & Nas, S., 2017. The impact of a single

tuition fee on the quality of education services.

Management of Journal of Management Education,

4(2), 184-194.

Surtiati, Siregar, H., & Andati, T., 2017. Cash flow analysis

related to single tuition policies at state law institutions.

Journal of Business and Management Applications,

3(2), 222-235. doi: 10.17358/JABM.3.2.222

Tang, T. L.-P., Tang, D. S.-H., & Tang, C. S.-Y., 2004.

College tuition and perceptions of private university

quality. International Journal of Educational

Management, 18(5), 304-316. doi: http://dx.doi.org/10.

1108/09513540410543457

Vizioli, R., & Kaminski, P. C., 2017. Problem definition as

a stimulus to the creative process: analysis of a

classroom exercise. Journal of Technology and Science

Education, 7(3), 274-290. doi: http://dx.doi.org/10.

3926/jotse.175

Wanjala, G., & Ali, K. H., 2017. Impact of subsidized fees

on students’ access to quality education in public

secondary schools in wajir county, kenya. International

Journal of Education and Research, 5(7), 247-262.

Wardi, Y., Abror, & Trinanda, O., 2018. The marketing of

higher education: managing student loyalty based on

tuition fee policy and service quality. Business and

Management Journal, 19(2), 101-108.

Wibawa, A. P., Kurniawan, A. C., Murti, D. M.,

Adiperkasa, R. P., Putra, S. M., Kurniawan, S. A., &

ICEIS 2021 - 23rd International Conference on Enterprise Information Systems

286

Nugraha, Y. R., 2019. Naïve Bayes Classifier for

Journal Quartile Classification. International Journal of

Recent Contributions from Engineering, Sciences, & IT,

7(2), 91-99. doi: https://doi.org/10.3991/ijes.v7i2.10659

Zhang, W., & Gao, F., 2011. An Improvement to Naive

Bayes for Text Classification. Advanced in Control

Engineering and Informaion Science, 15, 2160-2164.

doi: 10.1016/j.proeng.2011.08.404

APPENDIX

Appendix 1: List of STF Parameters / Criteria.



10.000.000

3. Mother's income

 500.000

 750.000

 1.000.000

 1.250.000

 1.500.000

 2.000.000



2.500.000

 3.000.000

 4.000.000

 5.000.000

 6.0000.000

 7.000.000

 8.000.000

 9.000.000

 10.000.000

4. Electrical

owe

 450kwh

 900kwh

 1300kwh

 2200kwh

 >2200kwh

5. Number of families



>6

 6

 4

 3

 2

6. Number of de

endents

 >4

 4

 3

 2

 1

STF Parameters / Criteria STF Groups

1. Parents' Income

 0‐500.000

 500‐1.000.000

 1.000.0000‐1.500.000

 1.500.000‐2.000.000

 2.000.000‐2.500.000

 3.000.000‐4.000.000

 4.500.000‐5.000.000

 5.000.000‐6.000.000

 6.500.000‐7.000.000

 7.500.000‐10.000.000

 10.000.000‐15.000.000

2. Father's income

 500.000

 750.000

 1.000.000

 1.250.000

 1.500.000

 2.000.000

 2.500.000

 3.000.000

 4.000.000

 5.000.000

 6.0000.000

 7.000.000

 8.000.000

 9.000.000

 10.000.000

3. Mother's income

 500.000

 750.000

 1.000.000

 1.250.000

 1.500.000

 2.000.000

 2.500.000

 3.000.000

 4.000.000

 5.000.000

 6.0000.000

 7.000.000

 8.000.000

 9.000.000

 10.000.000

Estimated Costs for Single Tuition Fee (STF) using Naïve Bayes Method

287

Appendix 2: Data used as calculation experiments.

No Parents' Incomes

Father's

income

Mother's

income

Electrical

power

Number of

families

Number of

dependents

Indonesia

Smart Card

STF Group

1 2.500.000 - 3.000.000 - 2.000.000 1300 kwh 3 1 None Category III

2 500.00 - 1.000.000 6.000.000 - 900 kwh 5 2 None Category V

3 2.000.000 - 2.500.000 2.500.000 - 900 kwh 5 4 None Category III

4 1.500.000 - 2.000.000 2.000.000 - 900 kwh 4 3 None Category III

5 500.000 - 1.000.000 - - - - - None Category IV

6 500.000 - 1.000.000 - 1.000.000 450 kwh 5 >4 None Category II

7 500.000 - 1.000.000 500.000 - 450 kwh >6 3 None Category I

8 0 - 500.000 - 500.000 450 kwh 6 3 None Category I

9 0 - 500.000 1.000.000 - 450 kwh 4 3 None Category III

10 4.000.000 - 5.000.000 3.000.000 - 900 kwh 6 >4 None Category IV

11 7.500.000 - 10.000.000 - 500.000 900 kwh 3 2 None Category V

12 500.000 - 1.000.000 - - - - - None Category IV

13 500.000 - 1.000.000 - - - - - None Category IV

14 2.000.000 - 2.500.000 1.500.000 - 900 kwh 5 3 None Category III

15 500.000 - 1.000.000 1.000.000 - 450 kwh 3 2 None Category III

16 1.500.000 - 2.000.000 - - - - - None Category IV

17 1.000.000 - 1.500.000 - - 900 kwh >6 4 None Category IV

18 500.000 - 1.000.000 - - - - - None Category IV

19 1.000.000 - 1.500.000 - - - - - None Category IV

20 1.500.000 - 2.000.000 1.500.000 - 900 kwh 6 4 None Category III

21 1.500.000 - 2.000.000 1.500.000 - 900 kwh 3 1 None Category III

22 0 - 500.000 - 1.500.000 1300 kwh 3 2 None Category III

23 4.000.000 - 5.000.000 4.000.000 4.000.000 900 kwh 4 2 None Category V

24 1.000.000 - 1.500.000 1.500.000 - 450 kwh 4 3 None Category III

25 500.000 - 1.000.000 1.000.000 - 900 kwh 4 2 None Category III

26 2.500.000 - 3.000.000 3.000.000 - 450 kwh 4 3 None Category IV

27 1.500.000 - 2.000.000 - - - - - None Category IV

28 4.000.000 - 5.000.000 5.000.000 - 900 kwh 5 4 None Category V

29 0 - 500.000 - - - - - None Category IV

30 3.000.000 - 4.000.000 - 4.000.000 900 kwh 5 3 None Category IV

31 1.500.000 - 2.000.000 - - - - - None Category IV

32 0 - 500.000 - - - - - None Category IV

33 1.000.000 - 1.500.000 - - - - - None Category IV

34 1.500.000 - 2.000.000 - - - - - None Category IV

35 1.000.000 - 1.500.000 - - - - - None Category IV

36 500.000 - 1.000.000 - - - - - None Category IV

37 500.000 - 1.000.000 1.000.000 - 900 kwh 5 1 None Category III

38 0 - 500.000 750.000 - 450 kwh 6 2 Yes Category I

39 0 - 500.000 - - 450 kwh 3 1 Yes Category II

40 500.000 - 1.000.000 - 500.000 450 kwh 4 3 None Category I

400 0 - 500.000 - - 1300 kwh 4 3 None Category III

ICEIS 2021 - 23rd International Conference on Enterprise Information Systems

288

Appendix 3: Calculation Steps using Naive Bayes.

1. The first step is counting the number of classes/ labels

- P (Y = Group 1) = 4/40 "The amount of group 1 data in the

training data is divided by the total amount of data"

- P (Y = Group 2) = 2/40 "The amount of group 2 data in the

training data is divided by the total amount of data"

- P (Y = Group 3) = 12/40 "Total group 3 data in the training

data is divided by the total amount of data"

- P (Y = Group 4) = 18/40 "Amount of group 4 data in the

training data is divided into the total amount of data"

- P (Y = Group 5) = 4/40 "Amount of group 5 data in the

training data is divided into the total amount of data"

2. Second step is to count the same number of cases with the same

class

A1. P (Parents’ Income = 2,000,000-2,500,000 | Y = Group 1)

= 0/4

B1. P (Parents’ Income = 2,000,000-2,500,000 | Y = Group 2)

= 0/2

C1. P (Parents’ Income = 2,000,000-2,500,000 | Y = Group 3)

= 2/12

D1 P (Parents’ Income = 2,000,000-2,500,000 | Y = Group 4)

= 0/18

E1. P (Parents’ Income = 2,000,000-2,500,000 | Y = Group 5)

= 0/4

A2. P (father's income = 2,500,000 | Y = Group 1) = 0/4

B2. P (father's income = 2,500,000 | Y = group 2) = 0/2

C2 P (father's income = 2,500,000 | Y = group 3) = 1/12

D2 P (father's income = 2,500,000 | Y = group 4) = 0/18

E2 P (father's income = 2,500,000 | Y = group 5) = 0/4

A3. P (Mother’s income = 0 | Y = group 1) = 2/4

B3. P (Mother’s income = 0 | Y = group 2) = ½

C3. P (Mother’s income = 0 | Y = group 3) = 10/12

D3. P (Mother’s income = 0 | Y = group 4) = 16/18

E3. P (Mother’s income = 0 | Y = group 5) = 2/4

A4. P (Electrical power = 450kwh | Y = Group 1) = 4/4

B4. P (Electrical power = 450kwh | Y = Group 2) = 2/2

C4. P (Electrical power = 450kwh | Y = Group 3) = 3/12

D4. P (Electric power = 450kwh | Y = Group 4) = 1/18

E4. P (Electrical power = 450kwh | Y = Group 5) = 0/4

A5. P (Number of Families = 5 | Y = Group 1) = 0/4

B5. P (Number of Families = 5 | Y = Group 2) = ½

C5 P (Number of Families = 5 | Y = Group 3) = 3/12

D5 P (Number of Families = 5 | Y = Group 4) = 1/18

E5 P (Number of Families = 5 | Y = Group 5) = 2/4

A6. P (Number of Dependents = 4 | Y = Group 1) = 0/4

B6. P (Number of Dependents = 4 | Y = Group 1) = 0/2

C6. P (Number of Dependents = 4 | Y = Group 1) = 2/12

D6. P (Number of Dependents = 4 | Y = Group 1) = 1/18

E6. P (Number of Dependents = 4 | Y = Group 1) = ¼

A7. P (Smart Indonesia Card = None | Y = Group 1) = ¾

B7. P (Smart Indonesia Card = None | Y = Group 2) = ½

C7. P (Smart Indonesia Card = None | Y = Group 3) = 12/12

D7. P (Smart Indonesia Card = None | Y = Group 4) = 18/18

E7. P (Smart Indonesia Card = None | Y = Group 5) = 4/4

3. Thirdly, multiply all the results of Group 1, 2, 3, 4, and 5

variables

A. -A1*A2*A3*A4*A5*A6*A7=

0/4*0/4*2/4*4/4*0/4*0/4*3/4 = 0

B. -B1*B2*B3*B4*B5*B6*B7=0/2*

0/2*1/2*2/2*1/2*0/2*1/2=0

C. -C1*C2*C3*C4*C5*C6*C7=

2/12*1/12*10/12*3/12*3/12*2/12*12 =0.005

D. -D1*D2*D3*D4*D5*D6*D7=

0/18*0/18*16/18*1/18*1/18*1/18*18/18= 0

E. -E1*E2*E3*E4*E5*E6*E7=

0/4*0/4*2/4*0/4*2/4*1/4*4/4= 0

4. Then compare and identify the highest probability value with the

results seen from the C value which is 0.005. This means it can

be categorized in group 3.

Estimated Costs for Single Tuition Fee (STF) using Naïve Bayes Method

289