<![CDATA[Savings and Loan Cooperatives often face the challenge of defaulted loans, which pose risks to financial stability and member trust. This study aims to compare the performance of the K-Nearest Neighbors (KNN) and Naive Bayes algorithms in classifying loan eligibility using the CRISP-DM (Cross-Industry Standard Process for Data Mining) approach. A case study was conducted on member financing data to identify a more accurate classification model to minimize loan defaults. The CRISP-DM methodology encompasses business understanding, data analysis, data preparation, modeling, evaluation, and deployment. The results show that KNN achieved the highest accuracy rate of 92.86%, while Naive Bayes only reached 85.71%. Additionally, KNN outperformed Naive Bayes in terms of precision and recall. Thus, KNN was selected as the optimal model to assist cooperatives in predicting loan eligibility. The implementation of this model is expected to improve financing efficiency, reduce default risks, and strengthen data-driven decision-making in cooperatives.]]>
Copyrights © 2024
