A Self-Organizing Maps–Based Approach for Home Ownership Loan Applicant Segmentation

The dataset used in this study comprises five key input variables: age, occupation, income level, loan amount, and credit history, which represent the core attributes commonly considered in loan evaluation processes. The SOM algorithm is applied to identify underlying similarity patterns among applicants without relying on predefined weights or decision rules. The clustering results reveal the formation of four distinct clusters, each representing different applicant profiles and characteristics related to credit eligibility and risk.