Application Research On Rough Sets In Credit Risk Measurement Of Commercial Bank

Commercial Banks play an important role in modern financial system and serve the social capital transaction due to the huge amount of capital and massive customers. Currently, Loan business is one of the most significant business in Commercial Banks, which also makes a great contribution for the profits.Under the background of strong policy support and huge market demand, credit volume is rising in recent years, credit risk is also rising.In fact, Commercial banks will fall into crisis and even bankruptcy if credit risk doesn’t get an appropriate treatment.The situation will become worse under the domino effect and finally the entire financial system will collapse.Therefore, credit risk management and control in the lending process not only became an important part of commercial bank management, but also related to the development of China’s entire financial market.Credit risk management and control systems cover risk measurement, response, monitoring etc. Actually, credit risk response and monitoring won’t succeed if credit risk doesn’t get a scientific and reasonable measurement.Commercial banks used to measure borrower’s financial factors and ignored the impact of the non-financial factors.Firstly, we summarize the credit risk measurement index on the basis of the previous studies and analyze the primary index by Delphi method, which can help us to create a scientific and reasonable index system and make a good preparation for our measurement.Then we create credit risk measurement model by the Rough set, in this part we give detailed instructions for each step of the chosen method. For attribute reduction we describe common reduction algorithm and introduce GA(Genetic Algorithm) due to the large amount of data of the credit risk measurement, in this part, we try to use the improved algorithm which can speed up the convergence and save running time relative to the general genetic algorithm, which can be applied for the attribute reduction of credit risk measurement. Finally we choose our listed company data to analyze, the empirical analysis reflects the credit risk measurement under rough set is well and can provide a reference for making the credit decision process to some extent.