Reputation-based Learning Nodes Selection Algorithm for Decentralized Federated Learning
Due to the increasing risk of privacy leakage,the collected data usually contains a large amount of privacy information,data collectors are reluctant to share their private data,which leads to result in"data silos".Federated learning enables data sharing without leaving the local area,but there are still some problems on sharing among multiple data.On the one hand,the centralized processing of central server suffers from expensive cost and single point of failure.On the other hand,for multi-institutional data sha-ring,model training might be affected by participating nodes mixed with malicious nodes,which leads to data privacy leakage.There-fore,Based on above analysis,a new distributed federated learning architecture is proposed to combine blockchain and federated learn-ing,realize the efficient node selection and communication,and it enables direct communication between participation nodes instead of relying on central server.Based on the proposed architecture,a reputation-based node selection algorithm scheme(RBLNS)is pro-posed to screen the participating nodes,and ensure the privacy and security of the participating nodes.The experimental results show that the RBLNS significantly improves the test performance of the model.
万方数据
维普
