Farhan Setya Dhitama
Fakultas Ilmu Komputer, Universitas Brawijaya

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

Penentuan Kelayakan Debitur Menggunakan Metode Decision Tree C4.5 Dan Oversampling Adaptive Synthetic (ADASYN) Farhan Setya Dhitama; Fitra Abdurrachman Bachtiar
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 4 No 10 (2020): Oktober 2020
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Credit is an activity or service that cannot be separated from life in the current era. Credit can also be interpreted as a loan of money, goods or services that have a limited time agreement and may include guarantees or not. Nowadays there are many companies in Indonesia that provide credit services. One of the challenges for companies engaged in credit provision is the credit that is delinquent. Less precisely judgment at the beginning debtors want to apply for credit being the cause of the credit that delinquent itself. This research aims to analyze and determine the feasibility decision of prospective debtor to receive credit from the credit provider bank in Lamongan. In the decision making system of credit eligibility, the method of decision Tree C 4.5 was used to classify into accepted classes or rejection potential debtors and also use Adaptive Synthetic (ADASYN) methods to perform oversampling processes on minority classes, as highly data that has been rejected is unbalanced in number with data that received credit decisions. The study uses the Decision Tree C 4.5 method as the debtor feasibility technique and the ADASYN method as an oversampling technique on the data that has the minority class. The features of the data used are Character, Capital, Capacity, Condition, Collateral, Age, and Dependents. The data to be used for classification calculations will be normalised using the Z-Score equation so that the data spread is not too wide. This research successfully develops a system that can classify debtor's eligibility using the Decision Tree C 4.5 and Adaptive Synthetic (ADASYN) methods for oversampling in the imbalance class. The test results show the best evaluation gained when the minor data sharing in training is 5 and in the testing amount of 2 and for the depth classification parameter of 1 and k is worth 3. Accuracy, Precision, Recall, and F-Measure obtained in this research is the Accuracy of getting 90%, Precision 100%, Recall worth 89%, and F-Measure is worth 94%.