IntroductioCredit risk is a determining factor in financial market, it is well known that one of aspects of Credit risk is loan application, especially, loan application is very important to the banks. The study of banks credit risk management under the new situation is directly related to the economic benefits and losses of banks. There are mainly two types of risks which are associated with the banks decision: if the applicant is a high quality one when it is likely to repay the loan to the back so in this situation, determining not approving the loan to the client is a loss of benefit, and if the applicant has a bad credit risk when it is unlikely to repay the loan, then approving the loan results a loss too. Thus, there is a requirement for the bank to design a rule which balance the approval of the loan to client and not approval to minimize the total loss. In this project, the goal is try to find a most appropriate way to make decisions for banks to minimize the loss faced with lots of Loan applications, i.e., we aim to find models which could help banks make predictions of clients as bad or good as accuracy as possible.