Analyzing the impact of loan features on bank loan prediction using R andom F orest algorithm

Abstract Loans are a crucial source of income for the financial sector, but they also come with significant financial risks. The interest on loans constitutes a significant portion of a bank's assets. The demand for loans is growing worldwide, and organizations are devising efficient business strategies to attract more clients. Every day, a large number of people apply for loans for various reasons, but not all of them can be approved due to the risk of loan default. It is not uncommon for people to default on their loans, causing significant losses to banks. The purpose of this article is to determine whether to grant loans to specific individuals or organizations. The Random Forest Regressor model has been utilized to measure performance and identify suitable customers for loan approval. The model suggests that banks should not only target affluent clients but also consider other customer characteristics that are critical in credit granting and predicting loan default. The research examines various loan approval parameters such as gender, educational qualification, employment type, business type, loan term, and marital status. Additionally, the study analyzes the number of approved, drawn, and rejected loans, which provides valuable insights into loan approval and prediction.