Design of bank credit risk analysis system based on machine learning

Abstract

Credit business is one of the most important components in many businesses of a bank, and has a very important impact on the bank's income and development. The most important part of credit business is the assessment of credit risk. Accurate assessment can enable banks to increase returns with the lowest possible risk. Traditional credit risk assessment models will face difficulties in feature selection in the high-dimensional and sparse big data environment; in addition, the high noise in big data will also affect the evaluation effect of the model. In response to the above problems, this paper firstly analyzes the common credit risk control algorithms. Then, based on the framework of deep learning, a stack noise reduction self-encoding neural network algorithm is designed, which is applied to the problem of bank credit risk assessment. Through experimental demonstration and comparative analysis, the use of deep learning for credit risk assessment in the big data environment can better provide early warning of credit risks and reduce bank credit risks.