%A NIM.: 20106010007 Isnaini Fitriana Rahmawati
%O Noor Saif Muhammad Mussafi, S. Si., M.Sc., Ph.D., dan Muhamad Rashif Hilmi, S.Si., M.Sc.
%T KLASIFIKASI PENILAIAN KREDIT DENGAN METODE PEMBELAJARAN KERNEL
%X Credit scoring is a system used by financial institutions to assess the eligibility of loan applicants. The digital era has made machine learning approaches such as Support Vector Machine (SVM) offer effective alternatives for classifying credit risk. This study aims to describe the steps of SVM, evaluate the performance of the SVM algorithm using Linear, Polynomial, Radial Basis Function (RBF), and Sigmoid Kernels, and determine the best method for credit scoring classification. The dataset used is the German Credit Data from the UCI Machine Learning Repository, consisting of 1000 customer records with 21 independent variables and 1 dependent variable, namely “Risk”. The research was conducted through chi-square tests and binary logistic regression to examine the relationship between independent and dependent variabeles, resulting in 17 significant independent variables that were used for SVM testing. Four SVM kernels Linear, Polynomial, Radial Basis Function (RBF), and Sigmoid were applied. The results showed that the Linear kernel achieved an accuracy of 75.67%, Polynomial 72.67%, RBF 75.33%, and Sigmoid 69.67%. the best performance in this classification task was obtained using the Linear kernel.
%K Support Vector Machine (SVM), Penilaian Kredit, Kener Linear, Kernel Polynomial, Kernel RBF, Kernel Sigmoid, Akurasi
%D 2025
%I UIN SUNAN KALIJAGA YOGYAKARTA
%L digilib72904