Prediction model of home loan eligibility status

A home loan is known as a mortgage which is a type of loan used to purchase a home or real estate. Home loans are essential for many individuals who are looking to buy a home without paying the entire price upfront. Confronting a prevalent challenge involving a notable of 60% rejection rate among home loan applicants, this issue was identified during the pre-pandemic year. Therefore, this study aims to develop an optimized home loan eligibility status prediction model. The study systematically evaluated Decision Tree approaches, incorporating variable selection for enhanced precision in model comparison. The findings of this study found that loan amount, marital status, property area, and co-applicant income as pivotal factors influencing home loan eligibility status. Notably, the Decision Tree Entropy model emerges as the optimal model, achieving a remarkable average squared error of 19.71%. This model exhibits superior performance, evidenced by a high sensitivity of 84.16% and accuracy of 59.78% according to the confusion matrix. The results are hoped to provide valuable insights into improving house loan eligibility prediction models, which could improve the banking sector’s decision-making procedures and in turn, improve consumer satisfaction.