Financial risk prediction method and system based on block chain federal learning

Abstract

The invention discloses a financial risk prediction method and system based on block chain federated learning, and the method comprises the steps: A, training a financial risk prediction model based on LSTM by a participant locally according to the held data, and searching for model parameters; b, transmitting the parameters to a block chain network module, collecting model parameters of each user by a federal learning node, and storing the model parameters to each participating node in a transaction form; c, generating a transaction block; all the participating nodes collect model parameters and package the model parameters into blocks, all the participating nodes decide that block-out weights belong to by running a PoQ consensus mechanism, and the participating nodes obtaining the block-out weights are added into the block chain network; step D, selecting a coordination node, the coordination node being used for converging training results, an aggregation node aggregating model parameters according to records and results of the block chain network, updating the financial risk prediction model, issuing the updated financial risk prediction model to each participating node, and starting a new round of training learning; and E, setting a contribution incentive evaluation mechanism.

Description

technical field [0001] The present invention relates to the technical field of platform resource allocation, in particular to a financial risk prediction method and system based on blockchain federated learning. Background technique [0002] As a decentralized, immutable, and shared distributed ledger and database, blockchain has many advantages and is suitable for research related to federated learning. The blockchain is a decentralized distributed database. All nodes in the entire blockchain network have the same rights. It avoids the disadvantages of data leakage in the centralized system and strengthens privacy protection and data security. As an open system, the blockchain allows all data owners to join and update the parameters of the model. In addition, the information in the blockchain is non-tamperable and traceable, which to a certain extent avoids damage to the system by malicious data owners. Therefore, the application of blockchain in federated learning can en...

AI Technical Summary

Problems solved by technology

[0003] The data involved in the field of intelligent financial risk control based on machine learning is multifaceted, but due to the problems of data privacy, data islands, data security laws and regulations among various financial institutions and enterprises, the data cannot be fully integrated

To this end, the existing technology proposes a federated learning framework. Although the current federated learning technical framework can solve the problem of data islands in financial institutions to a certain extent, they all require a third-party server or central database to cooperate with all parties. How to determine the first Three parties and ensuring their security are difficult problems for financial institutions

Images