Research On Credit Risk Assessment Of Commercial Banks Based On BP Neural Network

At the beginning of the 21 st century,with the increasing competition in the global commercial banking industry,the business cycle of various industries has gradually shortened,resulting in credit resources becoming scarce resources.Foreign mainstream commercial banks have adjusted their business structure,and asset credit business has become a new growth point for profits.At present,all commercial banks in China regard credit asset business as their key business development area.In order to seize market share,various banks have begun to develop a credit review and approval management system that is in line with their own development assets business,accelerating the pace of information construction of bank credit management systems.With the development of informationization construction of commercial banks in China and the completion of the construction of various basic business system platforms,the commercial banking industry stores a large amount of business data.These stock business data usually contain a large number of related,invisible and corrective Risk information with potential mining value.However,some important decisions of bank credit practitioners do not come from these stock business or customer information data,but based on the employment intuition of the bank credit management department and the guidance of the marketing indicators,the massive information in the bank information database has become " Data Graves,how to effectively use this data has become a top priority.This paper adopts BP neural network algorithm,based on the author's understanding of the existing bank loan approval process,establishes a risk rating model based on BP neural network,and then conducts risk analysis on the listed company's financial indicators in the previous year,and gathers the market this year.The performance is assessed by credit rating,a process of giving nonlinear learning to the model is established,and finally the data information of another group of listed companies is used for testing.The impact on the credit risk assessment of listed companies comes from many aspects,including the consideration of financial factors from financial indicators,as well as the stimulation of non-financial factors from the complex external market,so the authors use BP neural network algorithm nonlinearity.The learning ability,combined with the evaluation method of credit rating of existing bank credit customers,analyzes and constructs a new bank credit evaluation system with learning ability.