Research on the Influencing Factors of Home Loan Approvals

. The purpose of this paper is to comprehensively discuss the factors that may impact the approval of housing loans using Binary Logit Regression and Random Forest. As housing loans have become more common, the increase in non-performing rates of housing loans in the last two years has led to stricter loan approvals and greater uncertainty for loan applicants. This study examines data from home loan lenders to create a model for predicting an applicant’s ability to obtain a home loan and to assist applicants in planning for the future. The dataset comprises 480 loan records and 12 variables. The model based on Binary Logit Regression passes the Likelihood Ratio Test with a final prediction accuracy of 81.64%, which is considered acceptable. The results indicate that the applicant’s region and credit status significantly affect loan approval. Through the Random Forest, it is found that in addition to credit history, the weights for monthly applicant income and loan term are also high. Overall, applicants can predict loan approval based on the degree of influence of these factors.