Tekinfo: Jurnal Ilmiah Teknik Industri dan Informasi http://ejurnal.setiabudi.ac.id/ojs/index.php/tekinfo <p style="text-align: justify;"><span lang="en"><strong>Tekinfo ( Jurnal Ilmiah Teknik Industri dan Informasi )</strong> merupakan jurnal yang dikelola oleh program studi S1 Teknik Industri, Fakultas Teknik, Universitas Setia Budi yang terbit setiap enam bulan sekali yaitu Bulan Mei dan Bulan November pada setiap tahunnya. Naskah yang kami terbitkan mencakup bidang ilmu Teknik Industri dan Teknologi Informasi. Kami terbuka bagi para pembaca dan peneliti untuk berkontribusi mengirimkan naskah penelitian yang mencakup bidang ilmu tersebut.<br></span></p> en-US jurnaltekinfo@gmail.com (Ida Giyanti) tekinfo@setiabudi.ac.id (Adhie) Fri, 30 Apr 2021 04:31:42 +0000 OJS http://blogs.law.harvard.edu/tech/rss 60 Implementation of Decision Tree Algorithm (J.48) to Predict Risk of Credit in BMT http://ejurnal.setiabudi.ac.id/ojs/index.php/tekinfo/article/view/904 <p><em>Credit is crucial in financial institutions that affects the growth and development of these institutions. Weak supervision and management in the process of extending credit to customers can lead to high non-performing loans. This problem occured in one of the financial institutions that provides credit to customers, namely BMT X. Data for 2019 showed that there were 600 applications for multipurpose loans. Of these, only about 76% showed good collectability. The condition of credit collectability that is not optimal causes BMT X to spend more to collect installments that must be paid by the debtor directly. This bad credit causes losses to the financial institution. Thus, in providing credit, BMT X must be smart in assessing customer’ feasibility. The purpose of this research is to design credit policies in order to minimize the prediction errors of customers with bad credit category. The technique used in this research is classification data mining with the J.48 algorithm. To measure the effectiveness of an attribute in classifying a data sample set, it is necessary to select the attribute that has the greatest information gain which will be placed at the root node. This research produces six rules with an accuracy level of 80,2% so as it can be used by BMT X to search customer’s feasibility to gain credit.</em></p> <p><strong><em>Keywords</em></strong><strong>: </strong><em>Algorithm J.48, data mining, decision tree, credit risk</em></p> Atik Febriani, Violita Anggraini ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc-sa/4.0 http://ejurnal.setiabudi.ac.id/ojs/index.php/tekinfo/article/view/904 Fri, 30 Apr 2021 04:29:51 +0000