Several savings and loan cooperatives (KSP) in Kediri City, Indonesia have used a website-based information system for increasing efficiency. However, the financial health of KSP in Kediri City remains low because there are many delays in credit payments and even bad credit occurs. Manual profiling for approving loan application causes bad decisions. The management needs a function to obtain recommendation for approving loan application and also for distributing the loan service to potential members automatically. The purpose of this research is to develop a mobile based application for loan approval recommendation and loan distribution utilizing Machine Learning (ML) in KSP and to study the performance of the model using Support Vector Machine (SVM) method. It adopted Waterfall Method including analysis, design, implementation, and testing for two purposes including SVM model development and Android based application development. The dataset experienced preprocess including data cleaning, label encoding, and normalization. It obtained amounts to 150 data for loan approval recommendation and 150 data for loan distribution. Implementation stage includes developing Android based application and Python based ML. The testing stage uses functional testing for Android application and K-Fold Cross Validation for ML performance. Android application has two users, the first is admin that can manage member, retrieve loan approval recommendation, and manage loan application, then the second is member that can retrieve loan distribution and apply loan. The performance of the ML using SVM includes the accuracy of loan approval recommendation reached 90%, while loan distribution reached 85%.
Nugroho et al. (Thu,) studied this question.