Development of Fuzzy Logic Application to Determine

1 downloads 0 Views 1MB Size Report
Bahasa Pemrograman Java. Jurnal. Teknologi Informasi & Pendidikan. Vol. 5. No. 1. Mulyanto, E., Sutojo, T., ... Kereta Api Transmisi, Vol.10. No.2,9-13.
THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

Development of Fuzzy Logic Application to Determine Credit Limit Based on Total Deposit, Income and Collateral Inputs (Case Study: XYZ Credit Union) 1*

Yulianti Paula Bria , Yusvina De P. Lengga

2

Department of Informatics Engineering, Widya Mandira Catholic University, Kupang City [email protected]

Abstract XYZ Credit Union is a union which has 2,259 members in which there are 386 members who could not repay the loan on time. It is caused by credit analysis factors that were not accurate and also caused by financial reports were careless because it was still done manually. Therefore, it was necessary to develop the application of fuzzy logic to determine a credit limit based on total deposit, income and collateral inputs that were expected to help resolve problems in the process of calculation in determining the credit limit quickly and accurately. The fuzzy inference method used was Mamdani method, while the programming language used was C # (C Sharp) with MySQL as database. Fuzzy logic application produces an application that can be used to help employees, especially in the credit department at XYZ Credit Union in determining the credit limit based on total deposit, income and collateral that can minimize losses on union. Key words: fuzzy logic, fuzzy inference, Mamdani, C Sharp, Credit Union

1. Introduction XYZ Credit Union is one of the financial institutions and business entities that raise funds from members in the form of deposit (savings) and subsequently distribute to members in the form of credit. The purpose of this union is to improve the welfare of members and their families and improve the welfare of the surrounding community through serve saving and credit easily and efficiently supported by the participation of its members.

in terms of whether approved or not the loan application by the credit applicant. This approach is done to solve the problem in this research. Based on the above problems, it will be made development of fuzzy logic application to determine credit limit based on total deposit (savings), income and collateral inputs which is expected to help resolve problems in the process of calculation in determining the credit limit quickly and accurately so there is no longer a problem credits.

XYZ Credit Union often have the problem that the amount of loans granted by union to members can not pay on time both basic loan and interest money are set so that it becomes problematic. At January 2015, XYZ Credit Union has 2,259 members where there are 386 members who can not pay on time. It is caused by credit analysis that are not accurate and financial reports that careless, so that the union suffered losses.

2. Literature Review Fuzzy logic is a logic dealing with the concept of truth in part, in which classical logic to claim that everything can be expressed in terms of binary (0 or 1). Fuzzy logic is considered able to map an input into an output without ignoring the existing factors. The advantages of fuzzy logic are conceptually simple and easy to understand, can tolerate data imprecises, and fuzzy logic is based on natural language (Mulyanto et.al., 2010). Fuzzy logic is one of the important aspects that influence decision making (James and Chad, 2006).

To overcome these problems, need to be made an application to handle the process of calculation in determining the credit limit by using fuzzy logic, so it can be minimize bad credit that harm to the union. Previous research conducted by Sry and Puput (2013), has successfully developed a decision support system to determine of credit limit and provide decision alternatives

Mardison (2012) conducted research about ―Decision Suppot System In Credit Disbursement Customers of the Bank by

120

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

Using Fuzzy Logic And Java Programming Language‖ that can deliver the results in credit disbursement customers quickly, efficiently and effectively.

namely an employee of, credit department of XYZ Credit Union. Application analysis phase generated use case diagram. Use case diagram is a logical model of data or processes are created to describe people who are interested and associated with the application that can be seen in figure 1.

Arifah (2013) also conducted research about Mamdani Fuzzy method application in determining total production that can estimate the number of batik tulis production in the Melati Mekar Mandiri company with truth reaches 91.75%.

Input Data Pinjaman

Display Rule Base

Bria and Saidjuna (2013) also conducted research about fuzzy logic that can help cattle farmers to determine weight gain of cattle based on water and food consumption inputs.

Mencetak Data Laporan Petugas Display Fungsi Keanggotaan

Search by NBA / Nama

Figure 1. Use Case Diagram

Departing from previous studies, this research develop fuzzy logic application to determine credit limit based on three inputs namely the amount of savings (total deposit), income and collateral with output in the form of credit limit, where the fuzzy inference method used is Mamdani and defuzzification method is centroid. This method can provide output or output that is more accurate and more in accordance with the pattern of problems to be solved and to be accepted by many people.

3.3 Application Design Phase In the design stage of the application, there are 3 main fuzzification process, the process of reasoning (evaluation rule in the rule base) and defuzzification. In the process will be determined fuzzification membership function (MF) of 3 input variables (amount of savings, income and collateral) and 1 output variable (big loan). Next will be the process of reasoning or fuzzy matching between the values of the rule base fuzzification process is used as a knowledge base. Rule base contains rules that form if ... then (IF ... THEN). Output produced in this section is fuzzy output. Decision-making technique used is the max-min. At max-min method, decisions are based on the rules of operation according to Mamdani. Furthermore fuzzy output is converted into numbers firmly in the process of using centroid defuzzification as the final result of the application that is used to determine the credit limit. A block diagram of fuzzy logic can be seen in figure 2.

3. Methods 3.1 Data Retrieval Phase Data retrieval is adapted to the needs of the application that is as follows: a. Observation Observation is conducted by direct intervention on the XYZ Credit Union, especially in the credit department to examine and inquire about problems in determining the credit limit. b. Literature study At this stage we collected and looked for data relating to credit or loan and references to fuzzy logic through books, journals and internet. c. Interview Interviews is conducted with relevant parties, ie employees of XYZ Credit Union, especially in the credit department in order to obtain information related to this research, especially the creation of the rule base.

Input Value Tegas Input MF

Fuzzification

Fuzzy Input

Reasoning

3.2 Application Analysis Phase The application is made for inputting data such as the amount of savings (total deposit), income and collateral and provide output in the form of large loans (credit limit), where the application has only one user role

Fuzzy Output

Output MF

Defuzzification

Output Value Tegas

121

Rule Base

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

Figure 2. Block diagram of fuzzy logic (Wahyudi, 2005).

3.3.1 Fuzzification Fuzzification is the process of changing the input system that has a firm value become linguistic variables using membership functions which is stored in the knowledge base. Membership function is a curve that shows the mapping of points of input data into membership values with interval of 0 to 1. The following explanation will show the membership functions of linguistic variables from the input variables such as total deposit, income and collateral as inputs and credit limit as an output.

Range

: 0 – 4 (Rp)

Value of small linguistic variable (Sedikit)

: 0, 0, 1, 2

Value of middle linguistic variable (sedang)

: 1, 2, 3

Value of large linguistic variable (besar)

: 2, 3, 4, 4

3.

Membership 1

Function

Tidal Layak

of

Layak

collateral

Sangat Layak

Derajat Keanggotaan µ[x]

1. Membership Function of total deposit 0 1

Rendah

1

Sedang

2

3

5

Domain Jaminan (Jumlah)

Besar

Figure 5. Membership function of collateral (domain jaminan)

Derajat Keanggotaan µ[x]

0 10

20

30

Domain Simpanan (Juta)

Figure 3. Membership function of total deposit (domain simpanan) Range

: 0 – 40 (Rp)

Value of small linguistic variable (rendah)

: 0, 0, 10, 20

Value of middle linguistic variable (sedang)

: 10, 20, 30

Value of large linguistic variable (besar)

: 20, 30, 40, 40

Range

: 0–5

Value of not proper linguistic variable (tidak layak)

: 0, 0, 1, 2

Value of proper linguistic variable (layak)

: 1, 2, 3

Value of very proper linguistic variable (sangat layak)

: 2, 3, 5 , 5

4. Membership Function of credit limit 1

Sedikit

Sedang

Tinggi

Derajat Keanggotaan µ[x]

0

2. Membership Function of income.

10

30

50

100

Domain Besar Pinjaman (Juta)

1

Sedikit

Sedang

Besar

Figure 6. Membership function of credit limit (domain besar pinjaman)

Derajat Keanggotaan µ[x]

0 1

2

Range

: 0 – 100 (Rp)

Value of small linguistic variable (sedikit)

: 0, 0, 10, 30

Value of middle linguistic variable (sedang)

: 10, 30, 50

Value of large linguistic variable (tinggi)

: 30, 50, 100, 100

3

Domain Penghasilan (Juta)

Figure 4. Membership function of income (domain penghasilan)

3.3.2 Reasoning Process Reasoning is the process of using fuzzy rules If-Then to transform fuzzy inputs into

122

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

fuzzy output. While the rule / knowledge base is a collection of knowledge or rules that necessary to achieve the objectives. Mechanism of fuzzy reasoning: match the fuzzification result with the rules which is exist in the knowledge base and display fuzzy operations to perform inference.

11. If deposit (simpanan) is middle (sedang) and income (penghasilan) is small (sedikit) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 12. If deposit (simpanan) is middle (sedang) and income (penghasilan) is small (sedikit) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is middle (sedang). 13. If deposit (simpanan) is middle (sedang) and income (penghasilan) is middle (sedang) and (collateral) jaminan is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 14. If deposit (simpanan) is middle (sedang) and income (penghasilan) is middle (sedang) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 15. If deposit (simpanan) is middle (sedang) and income (penghasilan) is middle (sedang) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 16. If deposit (simpanan) is middle (sedang) and income (penghasilan) is large (banyak) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 17. If deposit (simpanan) is middle (sedang) and income (penghasilan) is large (banyak) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 18. If deposit (simpanan) is middle (sedang) and income (penghasilan) is large (banyak) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 19. If deposit (simpanan) is large (besar) and income (penghasilan) is small (sedikit) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 20. If deposit (simpanan) is large (besar) and income (penghasilan) is small (sedikit) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is large (tinggi). 21. If deposit (simpanan) is large (besar) and income (penghasilan) is small (sedikit) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 22. If deposit (simpanan) is large (besar) and income (penghasilan) is middle (sedang) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit).

There are knowledge base of the application: 1. If deposit (simpanan) is small (rendah) and income (penghasilan) is small (sedikit) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 2. If deposit (simpanan) is small (rendah) and income (penghasilan) is small (sedikit) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 3. If deposit (simpanan) is small (rendah) and income (penghasilan) is small (sedikit) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is middle (sedang). 4. If deposit (simpanan) is small (rendah) and income (penghasilan) is middle (sedang) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 5. If deposit (simpanan) is small (rendah) and income (penghasilan) is middle (sedang) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 6. If deposit (simpanan) is small (rendah) and income (penghasilan) is middle (sedang) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is middle (sedang). 7. If deposit (simpanan) is small (rendah) and income (penghasilan) is large (banyak) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 8. If deposit (simpanan) is small (rendah) and income (penghasilan) is large (banyak) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 9. If deposit (simpanan) is small (rendah) and income (penghasilan) is large (banyak) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 10. If deposit (simpanan) is middle (sedang) and income (penghasilan) is small (sedikit) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit).

123

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

23. If deposit (simpanan) is large (besar) and income (penghasilan) is middle (sedang) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is large (tinggi). 24. If deposit (simpanan) is large (besar) and income (penghasilan) is middle (sedang) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 25. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 26. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is large (tinggi). 27. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi).

 z  z  n

z

z 1 n

j

j

  z j 

..................................................(1)

j 1

Description: µ(z) = aggregation output of membership function z = fuzzy output value 4. Implementation This stage implementing thn phase using C # Sharp programming language. 4. Discussion of Results 4.1 Process Page Views Process page views consist of 3 part. There are fuzzification, reasoning and defuzzification part. Process page views for the application of fuzzy logic can be seen in figure 8.

3.3.3 Defuzzification The output of the rule evaluation is fuzzy value and will be changed in the form of firm value in the process of defuzzification with the help of output membership function and centroid defuzzification method. Defuzzification is the process of changing the amount of fuzzy presented in the form of the fuzzy output associations with the membership function to regain the firm form. This is necessary because it is known as the true scope of the regulatory process is firm value. Defuzzification method used is the centroid method. Centroid method is also known as a method of COA (Center of Area) or method Center of Gravity. In this method, the output value of firm obtained based on the center of gravity of the yield curve decision-making process that can be illustrated in the figure below.

Figure 8. Process page views

Process page view is used for inputting data such as: type of loan, nba, nik, name, savings, income and collateral that will be processed and produce output results in the amount of loans form as the credit limit. 4.2 Rule Base Page Views Rule base page views for the application of fuzzy logic can be seen in figure 9.

Figure 7. Centroid method

The equation of the centroid method:

Figure 9. Rule base page views

124

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

On the rule base page views contains 27 rules which is generated from application of fuzzy logic.

2. The application of calculation in determining the credit limit using fuzzy logic method is able to provide the results of decisions quickly and more accurately. 6. References

4.3 Membership Function Page Views On page views membership functions for fuzzy logic applications can be seen in figure 10.

Arifah, E. D. (2013). Aplikasi Metode Fuzzy Mamdani Dalam Penentuan Jumlah Produksi. Bria, Y.P., Saidjuna, D. (2014). Development Of Fuzzy Logic Application To Determine Weight Gain Of Madura, Bali And Dairy Cattle Based On Water And Food Consumed. International Seminar on Scientific Issues and Trends (ISSIT). Bina Sarana Informatika Jakarta Timur. James, W., Chad, G. (2006). Fuzzy Logic Control for Robot Maze Traversal: an Undergraduate Case Study, São Paulo, Brazil. Mardison. (2012). Sistem Pendukung Keputusan Dalam Pencairan Kredit Nasabah Bank Dengan Menggunakan Logika Fuzzy Dan Bahasa Pemrograman Java. Jurnal Teknologi Informasi & Pendidikan. Vol. 5 No. 1. Mulyanto, E., Sutojo, T., Suhartono, V. (2010). Kecerdasan Buatan. Yogyakarta. Andi. Sri, S., Puput, Y. (2013). Sistem Pendukung Keputusan Penentuan Plafon Kredit Dengan Fuzzy MDAM (Multiple Attribute Decissio Making) Menggunakanan Metode SAW (Simple Additive Weightin). Skripsi. STMIK Duta Bangsa Surakarta : Surakarta. Wahyudi. (2005). Implementasi Fuzzy Logic Controller Pada Sistem Pengereman Kereta Api Transmisi, Vol.10. No.2,9-13.

Figure 10. Membership function page views

Membership function page views contains a function to display a degree of membership in the form of a triangle and trapezoid curve, variable names, variable range, variable type, membership function name, the type of membership function and parameters of total deposit/savings, income, collateral and credit limit. 4.4 Report Page Views Page Views report for the application of fuzzy logic can be seen in figure 11.

Figure 11. Report Page Views

On the report page views contains a function to look back on the input data as a whole or by date, month and year. 5. Conclusion The conclusion that can be drawn from the results of the analysis and testing has been done on the development of the application of fuzzy logic to determine a credit limit based on total deposit, income and collateral as follows: 1. The application is made by using fuzzy logic can determine credit limit with common types of loans, special loans and microloans based on total deposit, income and collateral, and generate output in the form of credit to be received by the members.

125

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

Using Fuzzy Logic And Java Programming Language‖ that can deliver the results in credit disbursement customers quickly, efficiently and effectively.

namely an employee of, credit department of XYZ Credit Union. Application analysis phase generated use case diagram. Use case diagram is a logical model of data or processes are created to describe people who are interested and associated with the application that can be seen in figure 1.

Arifah (2013) also conducted research about Mamdani Fuzzy method application in determining total production that can estimate the number of batik tulis production in the Melati Mekar Mandiri company with truth reaches 91.75%.

Input Data Pinjaman

Display Rule Base

Bria and Saidjuna (2013) also conducted research about fuzzy logic that can help cattle farmers to determine weight gain of cattle based on water and food consumption inputs.

Mencetak Data Laporan Petugas Display Fungsi Keanggotaan

Search by NBA / Nama

Figure 1. Use Case Diagram

Departing from previous studies, this research develop fuzzy logic application to determine credit limit based on three inputs namely the amount of savings (total deposit), income and collateral with output in the form of credit limit, where the fuzzy inference method used is Mamdani and defuzzification method is centroid. This method can provide output or output that is more accurate and more in accordance with the pattern of problems to be solved and to be accepted by many people.

3.3 Application Design Phase In the design stage of the application, there are 3 main fuzzification process, the process of reasoning (evaluation rule in the rule base) and defuzzification. In the process will be determined fuzzification membership function (MF) of 3 input variables (amount of savings, income and collateral) and 1 output variable (big loan). Next will be the process of reasoning or fuzzy matching between the values of the rule base fuzzification process is used as a knowledge base. Rule base contains rules that form if ... then (IF ... THEN). Output produced in this section is fuzzy output. Decision-making technique used is the max-min. At max-min method, decisions are based on the rules of operation according to Mamdani. Furthermore fuzzy output is converted into numbers firmly in the process of using centroid defuzzification as the final result of the application that is used to determine the credit limit. A block diagram of fuzzy logic can be seen in figure 2.

3. Methods 3.1 Data Retrieval Phase Data retrieval is adapted to the needs of the application that is as follows: a. Observation Observation is conducted by direct intervention on the XYZ Credit Union, especially in the credit department to examine and inquire about problems in determining the credit limit. b. Literature study At this stage we collected and looked for data relating to credit or loan and references to fuzzy logic through books, journals and internet. c. Interview Interviews is conducted with relevant parties, ie employees of XYZ Credit Union, especially in the credit department in order to obtain information related to this research, especially the creation of the rule base.

Input Value Tegas Input MF

Fuzzification

Fuzzy Input

Reasoning

3.2 Application Analysis Phase The application is made for inputting data such as the amount of savings (total deposit), income and collateral and provide output in the form of large loans (credit limit), where the application has only one user role

Fuzzy Output

Output MF

Defuzzification

Output Value Tegas

121

Rule Base

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

Figure 2. Block diagram of fuzzy logic (Wahyudi, 2005).

3.3.1 Fuzzification Fuzzification is the process of changing the input system that has a firm value become linguistic variables using membership functions which is stored in the knowledge base. Membership function is a curve that shows the mapping of points of input data into membership values with interval of 0 to 1. The following explanation will show the membership functions of linguistic variables from the input variables such as total deposit, income and collateral as inputs and credit limit as an output.

Range

: 0 – 4 (Rp)

Value of small linguistic variable (Sedikit)

: 0, 0, 1, 2

Value of middle linguistic variable (sedang)

: 1, 2, 3

Value of large linguistic variable (besar)

: 2, 3, 4, 4

3.

Membership 1

Function

Tidal Layak

of

Layak

collateral

Sangat Layak

Derajat Keanggotaan µ[x]

1. Membership Function of total deposit 0 1

Rendah

1

Sedang

2

3

5

Domain Jaminan (Jumlah)

Besar

Figure 5. Membership function of collateral (domain jaminan)

Derajat Keanggotaan µ[x]

0 10

20

30

Domain Simpanan (Juta)

Figure 3. Membership function of total deposit (domain simpanan) Range

: 0 – 40 (Rp)

Value of small linguistic variable (rendah)

: 0, 0, 10, 20

Value of middle linguistic variable (sedang)

: 10, 20, 30

Value of large linguistic variable (besar)

: 20, 30, 40, 40

Range

: 0–5

Value of not proper linguistic variable (tidak layak)

: 0, 0, 1, 2

Value of proper linguistic variable (layak)

: 1, 2, 3

Value of very proper linguistic variable (sangat layak)

: 2, 3, 5 , 5

4. Membership Function of credit limit 1

Sedikit

Sedang

Tinggi

Derajat Keanggotaan µ[x]

0

2. Membership Function of income.

10

30

50

100

Domain Besar Pinjaman (Juta)

1

Sedikit

Sedang

Besar

Figure 6. Membership function of credit limit (domain besar pinjaman)

Derajat Keanggotaan µ[x]

0 1

2

Range

: 0 – 100 (Rp)

Value of small linguistic variable (sedikit)

: 0, 0, 10, 30

Value of middle linguistic variable (sedang)

: 10, 30, 50

Value of large linguistic variable (tinggi)

: 30, 50, 100, 100

3

Domain Penghasilan (Juta)

Figure 4. Membership function of income (domain penghasilan)

3.3.2 Reasoning Process Reasoning is the process of using fuzzy rules If-Then to transform fuzzy inputs into

122

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

fuzzy output. While the rule / knowledge base is a collection of knowledge or rules that necessary to achieve the objectives. Mechanism of fuzzy reasoning: match the fuzzification result with the rules which is exist in the knowledge base and display fuzzy operations to perform inference.

11. If deposit (simpanan) is middle (sedang) and income (penghasilan) is small (sedikit) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 12. If deposit (simpanan) is middle (sedang) and income (penghasilan) is small (sedikit) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is middle (sedang). 13. If deposit (simpanan) is middle (sedang) and income (penghasilan) is middle (sedang) and (collateral) jaminan is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 14. If deposit (simpanan) is middle (sedang) and income (penghasilan) is middle (sedang) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 15. If deposit (simpanan) is middle (sedang) and income (penghasilan) is middle (sedang) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 16. If deposit (simpanan) is middle (sedang) and income (penghasilan) is large (banyak) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 17. If deposit (simpanan) is middle (sedang) and income (penghasilan) is large (banyak) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 18. If deposit (simpanan) is middle (sedang) and income (penghasilan) is large (banyak) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 19. If deposit (simpanan) is large (besar) and income (penghasilan) is small (sedikit) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 20. If deposit (simpanan) is large (besar) and income (penghasilan) is small (sedikit) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is large (tinggi). 21. If deposit (simpanan) is large (besar) and income (penghasilan) is small (sedikit) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 22. If deposit (simpanan) is large (besar) and income (penghasilan) is middle (sedang) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit).

There are knowledge base of the application: 1. If deposit (simpanan) is small (rendah) and income (penghasilan) is small (sedikit) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 2. If deposit (simpanan) is small (rendah) and income (penghasilan) is small (sedikit) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 3. If deposit (simpanan) is small (rendah) and income (penghasilan) is small (sedikit) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is middle (sedang). 4. If deposit (simpanan) is small (rendah) and income (penghasilan) is middle (sedang) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 5. If deposit (simpanan) is small (rendah) and income (penghasilan) is middle (sedang) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 6. If deposit (simpanan) is small (rendah) and income (penghasilan) is middle (sedang) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is middle (sedang). 7. If deposit (simpanan) is small (rendah) and income (penghasilan) is large (banyak) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 8. If deposit (simpanan) is small (rendah) and income (penghasilan) is large (banyak) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 9. If deposit (simpanan) is small (rendah) and income (penghasilan) is large (banyak) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 10. If deposit (simpanan) is middle (sedang) and income (penghasilan) is small (sedikit) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit).

123

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

23. If deposit (simpanan) is large (besar) and income (penghasilan) is middle (sedang) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is large (tinggi). 24. If deposit (simpanan) is large (besar) and income (penghasilan) is middle (sedang) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 25. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 26. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is large (tinggi). 27. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi).

 z  z  n

z

z 1 n

j

j

  z j 

..................................................(1)

j 1

Description: µ(z) = aggregation output of membership function z = fuzzy output value 4. Implementation This stage implementing thn phase using C # Sharp programming language. 4. Discussion of Results 4.1 Process Page Views Process page views consist of 3 part. There are fuzzification, reasoning and defuzzification part. Process page views for the application of fuzzy logic can be seen in figure 8.

3.3.3 Defuzzification The output of the rule evaluation is fuzzy value and will be changed in the form of firm value in the process of defuzzification with the help of output membership function and centroid defuzzification method. Defuzzification is the process of changing the amount of fuzzy presented in the form of the fuzzy output associations with the membership function to regain the firm form. This is necessary because it is known as the true scope of the regulatory process is firm value. Defuzzification method used is the centroid method. Centroid method is also known as a method of COA (Center of Area) or method Center of Gravity. In this method, the output value of firm obtained based on the center of gravity of the yield curve decision-making process that can be illustrated in the figure below.

Figure 8. Process page views

Process page view is used for inputting data such as: type of loan, nba, nik, name, savings, income and collateral that will be processed and produce output results in the amount of loans form as the credit limit. 4.2 Rule Base Page Views Rule base page views for the application of fuzzy logic can be seen in figure 9.

Figure 7. Centroid method

The equation of the centroid method:

Figure 9. Rule base page views

124

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

On the rule base page views contains 27 rules which is generated from application of fuzzy logic.

2. The application of calculation in determining the credit limit using fuzzy logic method is able to provide the results of decisions quickly and more accurately. 6. References

4.3 Membership Function Page Views On page views membership functions for fuzzy logic applications can be seen in figure 10.

Arifah, E. D. (2013). Aplikasi Metode Fuzzy Mamdani Dalam Penentuan Jumlah Produksi. Bria, Y.P., Saidjuna, D. (2014). Development Of Fuzzy Logic Application To Determine Weight Gain Of Madura, Bali And Dairy Cattle Based On Water And Food Consumed. International Seminar on Scientific Issues and Trends (ISSIT). Bina Sarana Informatika Jakarta Timur. James, W., Chad, G. (2006). Fuzzy Logic Control for Robot Maze Traversal: an Undergraduate Case Study, São Paulo, Brazil. Mardison. (2012). Sistem Pendukung Keputusan Dalam Pencairan Kredit Nasabah Bank Dengan Menggunakan Logika Fuzzy Dan Bahasa Pemrograman Java. Jurnal Teknologi Informasi & Pendidikan. Vol. 5 No. 1. Mulyanto, E., Sutojo, T., Suhartono, V. (2010). Kecerdasan Buatan. Yogyakarta. Andi. Sri, S., Puput, Y. (2013). Sistem Pendukung Keputusan Penentuan Plafon Kredit Dengan Fuzzy MDAM (Multiple Attribute Decissio Making) Menggunakanan Metode SAW (Simple Additive Weightin). Skripsi. STMIK Duta Bangsa Surakarta : Surakarta. Wahyudi. (2005). Implementasi Fuzzy Logic Controller Pada Sistem Pengereman Kereta Api Transmisi, Vol.10. No.2,9-13.

Figure 10. Membership function page views

Membership function page views contains a function to display a degree of membership in the form of a triangle and trapezoid curve, variable names, variable range, variable type, membership function name, the type of membership function and parameters of total deposit/savings, income, collateral and credit limit. 4.4 Report Page Views Page Views report for the application of fuzzy logic can be seen in figure 11.

Figure 11. Report Page Views

On the report page views contains a function to look back on the input data as a whole or by date, month and year. 5. Conclusion The conclusion that can be drawn from the results of the analysis and testing has been done on the development of the application of fuzzy logic to determine a credit limit based on total deposit, income and collateral as follows: 1. The application is made by using fuzzy logic can determine credit limit with common types of loans, special loans and microloans based on total deposit, income and collateral, and generate output in the form of credit to be received by the members.

125

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

Figure 2. Block diagram of fuzzy logic (Wahyudi, 2005).

3.3.1 Fuzzification Fuzzification is the process of changing the input system that has a firm value become linguistic variables using membership functions which is stored in the knowledge base. Membership function is a curve that shows the mapping of points of input data into membership values with interval of 0 to 1. The following explanation will show the membership functions of linguistic variables from the input variables such as total deposit, income and collateral as inputs and credit limit as an output.

Range

: 0 – 4 (Rp)

Value of small linguistic variable (Sedikit)

: 0, 0, 1, 2

Value of middle linguistic variable (sedang)

: 1, 2, 3

Value of large linguistic variable (besar)

: 2, 3, 4, 4

3.

Membership 1

Function

Tidal Layak

of

Layak

collateral

Sangat Layak

Derajat Keanggotaan µ[x]

1. Membership Function of total deposit 0 1

Rendah

1

Sedang

2

3

5

Domain Jaminan (Jumlah)

Besar

Figure 5. Membership function of collateral (domain jaminan)

Derajat Keanggotaan µ[x]

0 10

20

30

Domain Simpanan (Juta)

Figure 3. Membership function of total deposit (domain simpanan) Range

: 0 – 40 (Rp)

Value of small linguistic variable (rendah)

: 0, 0, 10, 20

Value of middle linguistic variable (sedang)

: 10, 20, 30

Value of large linguistic variable (besar)

: 20, 30, 40, 40

Range

: 0–5

Value of not proper linguistic variable (tidak layak)

: 0, 0, 1, 2

Value of proper linguistic variable (layak)

: 1, 2, 3

Value of very proper linguistic variable (sangat layak)

: 2, 3, 5 , 5

4. Membership Function of credit limit 1

Sedikit

Sedang

Tinggi

Derajat Keanggotaan µ[x]

0

2. Membership Function of income.

10

30

50

100

Domain Besar Pinjaman (Juta)

1

Sedikit

Sedang

Besar

Figure 6. Membership function of credit limit (domain besar pinjaman)

Derajat Keanggotaan µ[x]

0 1

2

Range

: 0 – 100 (Rp)

Value of small linguistic variable (sedikit)

: 0, 0, 10, 30

Value of middle linguistic variable (sedang)

: 10, 30, 50

Value of large linguistic variable (tinggi)

: 30, 50, 100, 100

3

Domain Penghasilan (Juta)

Figure 4. Membership function of income (domain penghasilan)

3.3.2 Reasoning Process Reasoning is the process of using fuzzy rules If-Then to transform fuzzy inputs into

122

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

fuzzy output. While the rule / knowledge base is a collection of knowledge or rules that necessary to achieve the objectives. Mechanism of fuzzy reasoning: match the fuzzification result with the rules which is exist in the knowledge base and display fuzzy operations to perform inference.

11. If deposit (simpanan) is middle (sedang) and income (penghasilan) is small (sedikit) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 12. If deposit (simpanan) is middle (sedang) and income (penghasilan) is small (sedikit) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is middle (sedang). 13. If deposit (simpanan) is middle (sedang) and income (penghasilan) is middle (sedang) and (collateral) jaminan is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 14. If deposit (simpanan) is middle (sedang) and income (penghasilan) is middle (sedang) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 15. If deposit (simpanan) is middle (sedang) and income (penghasilan) is middle (sedang) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 16. If deposit (simpanan) is middle (sedang) and income (penghasilan) is large (banyak) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 17. If deposit (simpanan) is middle (sedang) and income (penghasilan) is large (banyak) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 18. If deposit (simpanan) is middle (sedang) and income (penghasilan) is large (banyak) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 19. If deposit (simpanan) is large (besar) and income (penghasilan) is small (sedikit) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 20. If deposit (simpanan) is large (besar) and income (penghasilan) is small (sedikit) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is large (tinggi). 21. If deposit (simpanan) is large (besar) and income (penghasilan) is small (sedikit) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 22. If deposit (simpanan) is large (besar) and income (penghasilan) is middle (sedang) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit).

There are knowledge base of the application: 1. If deposit (simpanan) is small (rendah) and income (penghasilan) is small (sedikit) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 2. If deposit (simpanan) is small (rendah) and income (penghasilan) is small (sedikit) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 3. If deposit (simpanan) is small (rendah) and income (penghasilan) is small (sedikit) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is middle (sedang). 4. If deposit (simpanan) is small (rendah) and income (penghasilan) is middle (sedang) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 5. If deposit (simpanan) is small (rendah) and income (penghasilan) is middle (sedang) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 6. If deposit (simpanan) is small (rendah) and income (penghasilan) is middle (sedang) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is middle (sedang). 7. If deposit (simpanan) is small (rendah) and income (penghasilan) is large (banyak) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 8. If deposit (simpanan) is small (rendah) and income (penghasilan) is large (banyak) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 9. If deposit (simpanan) is small (rendah) and income (penghasilan) is large (banyak) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 10. If deposit (simpanan) is middle (sedang) and income (penghasilan) is small (sedikit) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit).

123

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

23. If deposit (simpanan) is large (besar) and income (penghasilan) is middle (sedang) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is large (tinggi). 24. If deposit (simpanan) is large (besar) and income (penghasilan) is middle (sedang) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 25. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 26. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is large (tinggi). 27. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi).

 z  z  n

z

z 1 n

j

j

  z j 

..................................................(1)

j 1

Description: µ(z) = aggregation output of membership function z = fuzzy output value 4. Implementation This stage implementing thn phase using C # Sharp programming language. 4. Discussion of Results 4.1 Process Page Views Process page views consist of 3 part. There are fuzzification, reasoning and defuzzification part. Process page views for the application of fuzzy logic can be seen in figure 8.

3.3.3 Defuzzification The output of the rule evaluation is fuzzy value and will be changed in the form of firm value in the process of defuzzification with the help of output membership function and centroid defuzzification method. Defuzzification is the process of changing the amount of fuzzy presented in the form of the fuzzy output associations with the membership function to regain the firm form. This is necessary because it is known as the true scope of the regulatory process is firm value. Defuzzification method used is the centroid method. Centroid method is also known as a method of COA (Center of Area) or method Center of Gravity. In this method, the output value of firm obtained based on the center of gravity of the yield curve decision-making process that can be illustrated in the figure below.

Figure 8. Process page views

Process page view is used for inputting data such as: type of loan, nba, nik, name, savings, income and collateral that will be processed and produce output results in the amount of loans form as the credit limit. 4.2 Rule Base Page Views Rule base page views for the application of fuzzy logic can be seen in figure 9.

Figure 7. Centroid method

The equation of the centroid method:

Figure 9. Rule base page views

124

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

On the rule base page views contains 27 rules which is generated from application of fuzzy logic.

2. The application of calculation in determining the credit limit using fuzzy logic method is able to provide the results of decisions quickly and more accurately. 6. References

4.3 Membership Function Page Views On page views membership functions for fuzzy logic applications can be seen in figure 10.

Arifah, E. D. (2013). Aplikasi Metode Fuzzy Mamdani Dalam Penentuan Jumlah Produksi. Bria, Y.P., Saidjuna, D. (2014). Development Of Fuzzy Logic Application To Determine Weight Gain Of Madura, Bali And Dairy Cattle Based On Water And Food Consumed. International Seminar on Scientific Issues and Trends (ISSIT). Bina Sarana Informatika Jakarta Timur. James, W., Chad, G. (2006). Fuzzy Logic Control for Robot Maze Traversal: an Undergraduate Case Study, São Paulo, Brazil. Mardison. (2012). Sistem Pendukung Keputusan Dalam Pencairan Kredit Nasabah Bank Dengan Menggunakan Logika Fuzzy Dan Bahasa Pemrograman Java. Jurnal Teknologi Informasi & Pendidikan. Vol. 5 No. 1. Mulyanto, E., Sutojo, T., Suhartono, V. (2010). Kecerdasan Buatan. Yogyakarta. Andi. Sri, S., Puput, Y. (2013). Sistem Pendukung Keputusan Penentuan Plafon Kredit Dengan Fuzzy MDAM (Multiple Attribute Decissio Making) Menggunakanan Metode SAW (Simple Additive Weightin). Skripsi. STMIK Duta Bangsa Surakarta : Surakarta. Wahyudi. (2005). Implementasi Fuzzy Logic Controller Pada Sistem Pengereman Kereta Api Transmisi, Vol.10. No.2,9-13.

Figure 10. Membership function page views

Membership function page views contains a function to display a degree of membership in the form of a triangle and trapezoid curve, variable names, variable range, variable type, membership function name, the type of membership function and parameters of total deposit/savings, income, collateral and credit limit. 4.4 Report Page Views Page Views report for the application of fuzzy logic can be seen in figure 11.

Figure 11. Report Page Views

On the report page views contains a function to look back on the input data as a whole or by date, month and year. 5. Conclusion The conclusion that can be drawn from the results of the analysis and testing has been done on the development of the application of fuzzy logic to determine a credit limit based on total deposit, income and collateral as follows: 1. The application is made by using fuzzy logic can determine credit limit with common types of loans, special loans and microloans based on total deposit, income and collateral, and generate output in the form of credit to be received by the members.

125

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

fuzzy output. While the rule / knowledge base is a collection of knowledge or rules that necessary to achieve the objectives. Mechanism of fuzzy reasoning: match the fuzzification result with the rules which is exist in the knowledge base and display fuzzy operations to perform inference.

11. If deposit (simpanan) is middle (sedang) and income (penghasilan) is small (sedikit) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 12. If deposit (simpanan) is middle (sedang) and income (penghasilan) is small (sedikit) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is middle (sedang). 13. If deposit (simpanan) is middle (sedang) and income (penghasilan) is middle (sedang) and (collateral) jaminan is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 14. If deposit (simpanan) is middle (sedang) and income (penghasilan) is middle (sedang) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 15. If deposit (simpanan) is middle (sedang) and income (penghasilan) is middle (sedang) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 16. If deposit (simpanan) is middle (sedang) and income (penghasilan) is large (banyak) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 17. If deposit (simpanan) is middle (sedang) and income (penghasilan) is large (banyak) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 18. If deposit (simpanan) is middle (sedang) and income (penghasilan) is large (banyak) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 19. If deposit (simpanan) is large (besar) and income (penghasilan) is small (sedikit) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 20. If deposit (simpanan) is large (besar) and income (penghasilan) is small (sedikit) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is large (tinggi). 21. If deposit (simpanan) is large (besar) and income (penghasilan) is small (sedikit) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 22. If deposit (simpanan) is large (besar) and income (penghasilan) is middle (sedang) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit).

There are knowledge base of the application: 1. If deposit (simpanan) is small (rendah) and income (penghasilan) is small (sedikit) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 2. If deposit (simpanan) is small (rendah) and income (penghasilan) is small (sedikit) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 3. If deposit (simpanan) is small (rendah) and income (penghasilan) is small (sedikit) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is middle (sedang). 4. If deposit (simpanan) is small (rendah) and income (penghasilan) is middle (sedang) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 5. If deposit (simpanan) is small (rendah) and income (penghasilan) is middle (sedang) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 6. If deposit (simpanan) is small (rendah) and income (penghasilan) is middle (sedang) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is middle (sedang). 7. If deposit (simpanan) is small (rendah) and income (penghasilan) is large (banyak) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 8. If deposit (simpanan) is small (rendah) and income (penghasilan) is large (banyak) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is middle (sedang). 9. If deposit (simpanan) is small (rendah) and income (penghasilan) is large (banyak) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 10. If deposit (simpanan) is middle (sedang) and income (penghasilan) is small (sedikit) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit).

123

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

23. If deposit (simpanan) is large (besar) and income (penghasilan) is middle (sedang) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is large (tinggi). 24. If deposit (simpanan) is large (besar) and income (penghasilan) is middle (sedang) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 25. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 26. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is large (tinggi). 27. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi).

 z  z  n

z

z 1 n

j

j

  z j 

..................................................(1)

j 1

Description: µ(z) = aggregation output of membership function z = fuzzy output value 4. Implementation This stage implementing thn phase using C # Sharp programming language. 4. Discussion of Results 4.1 Process Page Views Process page views consist of 3 part. There are fuzzification, reasoning and defuzzification part. Process page views for the application of fuzzy logic can be seen in figure 8.

3.3.3 Defuzzification The output of the rule evaluation is fuzzy value and will be changed in the form of firm value in the process of defuzzification with the help of output membership function and centroid defuzzification method. Defuzzification is the process of changing the amount of fuzzy presented in the form of the fuzzy output associations with the membership function to regain the firm form. This is necessary because it is known as the true scope of the regulatory process is firm value. Defuzzification method used is the centroid method. Centroid method is also known as a method of COA (Center of Area) or method Center of Gravity. In this method, the output value of firm obtained based on the center of gravity of the yield curve decision-making process that can be illustrated in the figure below.

Figure 8. Process page views

Process page view is used for inputting data such as: type of loan, nba, nik, name, savings, income and collateral that will be processed and produce output results in the amount of loans form as the credit limit. 4.2 Rule Base Page Views Rule base page views for the application of fuzzy logic can be seen in figure 9.

Figure 7. Centroid method

The equation of the centroid method:

Figure 9. Rule base page views

124

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

On the rule base page views contains 27 rules which is generated from application of fuzzy logic.

2. The application of calculation in determining the credit limit using fuzzy logic method is able to provide the results of decisions quickly and more accurately. 6. References

4.3 Membership Function Page Views On page views membership functions for fuzzy logic applications can be seen in figure 10.

Arifah, E. D. (2013). Aplikasi Metode Fuzzy Mamdani Dalam Penentuan Jumlah Produksi. Bria, Y.P., Saidjuna, D. (2014). Development Of Fuzzy Logic Application To Determine Weight Gain Of Madura, Bali And Dairy Cattle Based On Water And Food Consumed. International Seminar on Scientific Issues and Trends (ISSIT). Bina Sarana Informatika Jakarta Timur. James, W., Chad, G. (2006). Fuzzy Logic Control for Robot Maze Traversal: an Undergraduate Case Study, São Paulo, Brazil. Mardison. (2012). Sistem Pendukung Keputusan Dalam Pencairan Kredit Nasabah Bank Dengan Menggunakan Logika Fuzzy Dan Bahasa Pemrograman Java. Jurnal Teknologi Informasi & Pendidikan. Vol. 5 No. 1. Mulyanto, E., Sutojo, T., Suhartono, V. (2010). Kecerdasan Buatan. Yogyakarta. Andi. Sri, S., Puput, Y. (2013). Sistem Pendukung Keputusan Penentuan Plafon Kredit Dengan Fuzzy MDAM (Multiple Attribute Decissio Making) Menggunakanan Metode SAW (Simple Additive Weightin). Skripsi. STMIK Duta Bangsa Surakarta : Surakarta. Wahyudi. (2005). Implementasi Fuzzy Logic Controller Pada Sistem Pengereman Kereta Api Transmisi, Vol.10. No.2,9-13.

Figure 10. Membership function page views

Membership function page views contains a function to display a degree of membership in the form of a triangle and trapezoid curve, variable names, variable range, variable type, membership function name, the type of membership function and parameters of total deposit/savings, income, collateral and credit limit. 4.4 Report Page Views Page Views report for the application of fuzzy logic can be seen in figure 11.

Figure 11. Report Page Views

On the report page views contains a function to look back on the input data as a whole or by date, month and year. 5. Conclusion The conclusion that can be drawn from the results of the analysis and testing has been done on the development of the application of fuzzy logic to determine a credit limit based on total deposit, income and collateral as follows: 1. The application is made by using fuzzy logic can determine credit limit with common types of loans, special loans and microloans based on total deposit, income and collateral, and generate output in the form of credit to be received by the members.

125

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

23. If deposit (simpanan) is large (besar) and income (penghasilan) is middle (sedang) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is large (tinggi). 24. If deposit (simpanan) is large (besar) and income (penghasilan) is middle (sedang) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi). 25. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is not proper (tidak layak) Then credit limit (besar pinjaman) is small (sedikit). 26. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is proper (layak) Then credit limit (besar pinjaman) is large (tinggi). 27. If deposit (simpanan) is large (besar) and income (penghasilan) is large (banyak) and collateral (jaminan) is very proper (sangat layak) Then credit limit (besar pinjaman) is large (tinggi).

 z  z  n

z

z 1 n

j

j

  z j 

..................................................(1)

j 1

Description: µ(z) = aggregation output of membership function z = fuzzy output value 4. Implementation This stage implementing thn phase using C # Sharp programming language. 4. Discussion of Results 4.1 Process Page Views Process page views consist of 3 part. There are fuzzification, reasoning and defuzzification part. Process page views for the application of fuzzy logic can be seen in figure 8.

3.3.3 Defuzzification The output of the rule evaluation is fuzzy value and will be changed in the form of firm value in the process of defuzzification with the help of output membership function and centroid defuzzification method. Defuzzification is the process of changing the amount of fuzzy presented in the form of the fuzzy output associations with the membership function to regain the firm form. This is necessary because it is known as the true scope of the regulatory process is firm value. Defuzzification method used is the centroid method. Centroid method is also known as a method of COA (Center of Area) or method Center of Gravity. In this method, the output value of firm obtained based on the center of gravity of the yield curve decision-making process that can be illustrated in the figure below.

Figure 8. Process page views

Process page view is used for inputting data such as: type of loan, nba, nik, name, savings, income and collateral that will be processed and produce output results in the amount of loans form as the credit limit. 4.2 Rule Base Page Views Rule base page views for the application of fuzzy logic can be seen in figure 9.

Figure 7. Centroid method

The equation of the centroid method:

Figure 9. Rule base page views

124

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

On the rule base page views contains 27 rules which is generated from application of fuzzy logic.

2. The application of calculation in determining the credit limit using fuzzy logic method is able to provide the results of decisions quickly and more accurately. 6. References

4.3 Membership Function Page Views On page views membership functions for fuzzy logic applications can be seen in figure 10.

Arifah, E. D. (2013). Aplikasi Metode Fuzzy Mamdani Dalam Penentuan Jumlah Produksi. Bria, Y.P., Saidjuna, D. (2014). Development Of Fuzzy Logic Application To Determine Weight Gain Of Madura, Bali And Dairy Cattle Based On Water And Food Consumed. International Seminar on Scientific Issues and Trends (ISSIT). Bina Sarana Informatika Jakarta Timur. James, W., Chad, G. (2006). Fuzzy Logic Control for Robot Maze Traversal: an Undergraduate Case Study, São Paulo, Brazil. Mardison. (2012). Sistem Pendukung Keputusan Dalam Pencairan Kredit Nasabah Bank Dengan Menggunakan Logika Fuzzy Dan Bahasa Pemrograman Java. Jurnal Teknologi Informasi & Pendidikan. Vol. 5 No. 1. Mulyanto, E., Sutojo, T., Suhartono, V. (2010). Kecerdasan Buatan. Yogyakarta. Andi. Sri, S., Puput, Y. (2013). Sistem Pendukung Keputusan Penentuan Plafon Kredit Dengan Fuzzy MDAM (Multiple Attribute Decissio Making) Menggunakanan Metode SAW (Simple Additive Weightin). Skripsi. STMIK Duta Bangsa Surakarta : Surakarta. Wahyudi. (2005). Implementasi Fuzzy Logic Controller Pada Sistem Pengereman Kereta Api Transmisi, Vol.10. No.2,9-13.

Figure 10. Membership function page views

Membership function page views contains a function to display a degree of membership in the form of a triangle and trapezoid curve, variable names, variable range, variable type, membership function name, the type of membership function and parameters of total deposit/savings, income, collateral and credit limit. 4.4 Report Page Views Page Views report for the application of fuzzy logic can be seen in figure 11.

Figure 11. Report Page Views

On the report page views contains a function to look back on the input data as a whole or by date, month and year. 5. Conclusion The conclusion that can be drawn from the results of the analysis and testing has been done on the development of the application of fuzzy logic to determine a credit limit based on total deposit, income and collateral as follows: 1. The application is made by using fuzzy logic can determine credit limit with common types of loans, special loans and microloans based on total deposit, income and collateral, and generate output in the form of credit to be received by the members.

125

THE 1ST INTERNATIONAL CONFERENCE ON INNOVATIVE RESEARCH ACROSS DISCIPLINES 2015

On the rule base page views contains 27 rules which is generated from application of fuzzy logic.

2. The application of calculation in determining the credit limit using fuzzy logic method is able to provide the results of decisions quickly and more accurately. 6. References

4.3 Membership Function Page Views On page views membership functions for fuzzy logic applications can be seen in figure 10.

Arifah, E. D. (2013). Aplikasi Metode Fuzzy Mamdani Dalam Penentuan Jumlah Produksi. Bria, Y.P., Saidjuna, D. (2014). Development Of Fuzzy Logic Application To Determine Weight Gain Of Madura, Bali And Dairy Cattle Based On Water And Food Consumed. International Seminar on Scientific Issues and Trends (ISSIT). Bina Sarana Informatika Jakarta Timur. James, W., Chad, G. (2006). Fuzzy Logic Control for Robot Maze Traversal: an Undergraduate Case Study, São Paulo, Brazil. Mardison. (2012). Sistem Pendukung Keputusan Dalam Pencairan Kredit Nasabah Bank Dengan Menggunakan Logika Fuzzy Dan Bahasa Pemrograman Java. Jurnal Teknologi Informasi & Pendidikan. Vol. 5 No. 1. Mulyanto, E., Sutojo, T., Suhartono, V. (2010). Kecerdasan Buatan. Yogyakarta. Andi. Sri, S., Puput, Y. (2013). Sistem Pendukung Keputusan Penentuan Plafon Kredit Dengan Fuzzy MDAM (Multiple Attribute Decissio Making) Menggunakanan Metode SAW (Simple Additive Weightin). Skripsi. STMIK Duta Bangsa Surakarta : Surakarta. Wahyudi. (2005). Implementasi Fuzzy Logic Controller Pada Sistem Pengereman Kereta Api Transmisi, Vol.10. No.2,9-13.

Figure 10. Membership function page views

Membership function page views contains a function to display a degree of membership in the form of a triangle and trapezoid curve, variable names, variable range, variable type, membership function name, the type of membership function and parameters of total deposit/savings, income, collateral and credit limit. 4.4 Report Page Views Page Views report for the application of fuzzy logic can be seen in figure 11.

Figure 11. Report Page Views

On the report page views contains a function to look back on the input data as a whole or by date, month and year. 5. Conclusion The conclusion that can be drawn from the results of the analysis and testing has been done on the development of the application of fuzzy logic to determine a credit limit based on total deposit, income and collateral as follows: 1. The application is made by using fuzzy logic can determine credit limit with common types of loans, special loans and microloans based on total deposit, income and collateral, and generate output in the form of credit to be received by the members.

125