Enhancement in the banking region very huge customers are applying for different types of loans which is available in the all bank. But the bank has its own boundary assets which grant the permission for limited people. Loan approval is a very long and important step in bank organization. Banking sector need more precise predicting model for better accuracy. Predicting the credit customer is the very difficult task in bank sector. The predicting system should approve and rejects the loan application system. Loans are the core business for banks. Customer dataset is taken for identifying the key customer. The data mining technique are used for predicting the loans which containing high dimensional data. It contains some redundant and inappropriate attributes in the dataset. Machine learning techniques helps to predicting outcomes from huge amount of data. In this methodology it helps to focus on attributes and feature selection for identifying loans approval customer. In this proposed work two machine learning algorithms, Random Forest (RF) and Boruta Algorthim are applied to predict the key customer of the loan approval. This experimental result concludes that accuracy of Boruta Algorthim is better as compared to Random Forest algorithm. The social network analysis technique is also used to predict and to identify the key customer for further loan analysis.

Feature Selection Random Forest Boruta Social Network Analysis.