Melyana, Risti (2025) Prediksi Pembiayaan Mikro Dalam Meminimalisir Resiko Pembiayaan Bermasalah Di Bank Syariah Indonesia (BSI) Sudirman dengan Perbandingan Metode Algoritma K-Nearest Neighbors dan Decision Tree. Other thesis, Universitas Islam Riau.

Text skripsi_203510180_watermark.pdf - Submitted Version Restricted to Registered users only Download (2MB) | Request a copy

Abstract

icrofinance prediction to minimize the risk of problematic financing at Bank Syariah Indonesia (BSI) Sudirman Branch Office. Financing risk that is not managed properly can have a negative impact on the financial performance and reputation of Islamic banking. In this study, a comparison was made between two algorithm classifications, namely K-Nearest Neighbors (KNN) and Decision Tree (C4.5) which identify potential customers who have the potential to cause problematic financing risks. Historical microfinance data was used as 271 training data and 30 test data with a 10-fold cross-validation technique. The evaluation results showed varying levels of accuracy between the two methods. Based on the test results, the method that showed the highest accuracy was considered more optimal in assisting the decision-making process for financing risk analysis at BSI. The results of the confusion matrix test on financing classification using the KNN and Decision methods can be applied to the accuracy value obtained in the KNN algorithm 73.33%, the Decision Tree algorithm 66.67%, the precision value of the KNN algorithm 73.27%, the Decision Tree algorithm 66.35%, the recall value of the K-NN algorithm 73.33%, the Decision Tree algorithm 66.67% and the F1-score value on K-NN 72.96% and the Decision Tree algorithm 66.20%. This research is expected to contribute to the development of decision support systems for Islamic financial institutions.

Actions (login required)

View Item