Perbandingan Algoritma KNN (K-Nearest Neighbors), Naïve Bayes, Dan SVM (Support Vector Machine) Untuk Klasifikasi Pemberian Pinjaman Nasabah

zairi saputra (STMIK Amik Riau)
H A Supahri (STMIK Amik Riau)
R Ismanizan (STMIK Amik Riau)
Rahmaddeni Rahmaddeni (STMIK Amik Riau)

Jurnal Ilmiah Sistem Informasi dan Teknik Informatika (JISTI)
Vol 7 No 1 (2024): Jurnal Ilmiah Sistem Informasi dan Teknik Informatika (JISTI)

AbstractThis journal examines the use of classification algorithms such as K-Nearest Neighbors (KNN), NaiveBayes, and Support Vector Machines (SVM) in providing loans to customers. This method is used toincrease the reliability and accuracy of the credit risk evaluation system. The experimentalmethodology involves a dataset containing variables related to credit history, income, and other riskfactors. The research results show that the KNN algorithm achieves a significant level of accuracy inidentifying customer risk profiles. On the other hand, Naive Bayes successfully handles data withdependencies between variables, and SVM provides consistent results in handling complex datasets.This research explores the benefits and drawbacks of each algorithm to help build a better decisionmaking system for customer lending.Keywords: KNN, Naïve Bayes, SVM, Customer Loans.