Distributed support vector machine based on blockchain and privacy protection and optimization method thereof

Abstract

The invention discloses a distributed support vector machine based on a blockchain and privacy protection and an optimization method thereof, and the method comprises the steps: building a distributed support vector machine system model with privacy protection based on the blockchain, and completing an interaction process between nodes according to a blockchain consensus PBFT protocol. By analyzing the calculation complexity of a local node in a training process and a consensus process, a calculation resource allocation optimization method based on consideration of energy consumption and system energy utilization rate is given. Simulation results show that the technical scheme and the model can provide privacy protection for the nodes and the model in training and consensus processes, and the utilization rate of total energy of the system and the performance of a distributed support vector machine learning process are improved by optimizing resources of each node and each step under the condition of energy consumption constraint.

Description

technical field [0001] The present invention relates to the related technical fields of privacy protection, data allocation and resource allocation in distributed machine learning. Specifically, it is a distributed support vector machine based on blockchain and privacy protection and its optimization method, and further involves the combination of vertical Distributed data nodes, the construction method of the alliance chain in the blockchain, the PBFT consensus mechanism and the calculation method of the privacy protection algorithm of the partial homomorphic encryption algorithm and the resource allocation optimization method. [0002] technical background [0003] In recent years, Internet data has grown day by day, so machine learning methods for processing data are widely used. Traditional machine learning methods aggregate all data into one machine or a data center, and a data analyst conducts centralized model training. However, due to privacy and security issues, in ...

AI Technical Summary

Problems solved by technology

However, due to privacy and security issues, in most cases, it is difficult for all learning participants to fully share data, so there is a very serious "data island" problem; and in the training process, traditional machine learning will cause privacy leakage problems; At the same time, since we are now in an era of Internet information explosion, it is very difficult for a single node to complete the acquisition of a complete database and subsequent training

Images