MULTINOMIAL NAIVE BAYES UNTUK KLASIFIKASI STATUS KREDIT MITRA BINAAN DI PT. ANGKASA PURA I PROGRAM KEMITRAAN
-
Meilani T Bunga(1)
Universitas Nusa Cendana
-
Bertha S Djahi(2*)
Universitas Nusa Cendana
-
Yelly Y Nabuasa(3)
Universitas Nusa Cendana
(*) Corresponding Author
Abstract
Status classification of partner acordiing to sector parimeter, loan disbursement, loan reimbursment, loan arrears, remaining loan and grace period is very important in anticipating the case in PT. Angkasa Pura I. Problematic credit is very unbeneficial for the PT. Angkasa Pura I because it will disturb the economy condition of a company and will affect the next small and medium enerprises (SME). To solve this, the reserch uses Multinominal Naive Bayes to method to classify the partners status in the PT. Angkasa Pura I according to the parimeter that is divided into 4 clases namely smooth class, less smooth class, doubted and jammed class. The process used was classification process where it calculated probability value and the atribute of the partner. The data used in this research is consisted of 148 that taken from 2012-2015. The final result, after the classification is done, the class probability value that was taken randomly is gained, with the resuld to system test with 5 times of testing data division that is taken randomly, it is gained the accuracy as big as 86,56%, precision is as big as 73%, recall is as big as 73% and F-1 Measure is as big as 73%.
Downloads
References
[2] Jiang, L., Wang, S., Li, C., Zhang, L., 2014, Structure extended multinomial naive bayes, doi:10.1007/s10115-014-0746-y, diakses 11 mei 2017.
[3] Raymond, McLeod, J., 2001, Sistem Informasi Edisi 7 Jilid 2, Prenhallindo, Jakarta.
[4] Saleh, A., 2014, Implementasi Metode Klasifikasi Naïve Bayes Dalam Memprediksi Besarnya Penggunaan Listrik Rumah Tangga, Yogyakarta.
Copyright (c) 2018 Jurnal Komputer dan Informatika
This work is licensed under a Creative Commons Attribution 4.0 International License.
The author submitting the manuscript must understand and agree that if accepted for publication, authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution (CC-BY) 4.0 License that allows others to share the work with an acknowledgment of the work’s authorship and initial publication in this journal.
