Financial credit data sharing method and device based on block chain and federal learning

Abstract

The invention provides a financial credit data sharing method and device based on a block chain and federated learning, and the method comprises the steps of carrying out the collection and feature analysis of credit data, and carrying out the federated learning, so as to train an evaluation model; selecting a plurality of moments to perform data truncation, performing iterative training and comparison on the truncated data through an evaluation model, verifying credit data increment, and generating an incremental data set; and fusing online learning and federated learning by using an incremental data set, updating credit features, and realizing dynamic updating of the user credit portrait under the supervision background. Compared with the prior art, the financial credit data sharing method and device based on the block chain and federal learning provided by the invention have the advantages that right confirmation evidence storage and transaction evidence storage of the financial credit data are realized, and continuous and dynamic individual and enterprise credit analysis evaluation and updating are achieved.

Description

technical field [0001] The invention relates to the technical field of block chains, in particular to a method and device for sharing financial credit data based on block chains and federated learning. Background technique [0002] At present, with the further development of China's market economy, my country is already in the era of Internet consumer finance. The new social consumption concept promotes the continuous expansion of the financial credit business of enterprises and individuals, and credit will become an important asset. Faced with a variety of corporate entities or individuals that need credit, how to comprehensively describe and evaluate their credit characteristics, and how to achieve reasonable risk identification and evaluation has become a core research issue for financial institutions. In this context, Internet big data has become an important supplement to financial credit. Traditional credit agencies represented by the People's Bank of China, Internet c...

AI Technical Summary

Problems solved by technology

Credit data is bound to specific real identities, which is directly related to the interests of enterprises and individuals, so the security requirements are more stringent, the cost of sharing and circulation in the traditional way is high, and there are risks of data modification and information leakage

The participation of third-party credit reporting agencies increases the risk of user credit data leakage and malicious use, and the data leakage cannot be monitored and traced

Distributed computing needs to be introduced into financial credit reporting, but there is a risk of data leakage in distributed computing

Images