Abstract | ||
---|---|---|
Movited by the modern phenomenon of distributed data collected by edge devices at scale, federated learning can use the large amounts of training data from diverse users for better representation and generalization. To improve flexibility and scalability, we propose a new federated optimization algorithm, named as Multi-index federated aggregation algorithm based on trusted verfication(TVFedmul). TVFedmul is optimized based on Fedavg algorithm, which overcomes a series of problems caused by the original aggregation algorithm, which only takes the single index of data quantity as a reference factor to measure the aggregation weight of each client. The improved aggregation algorithm is based on multi-index measurement, which can reflect the comprehensive ability of clients more comprehensively, so as to make overall judgment. Further, we introduces hyperparameter alpha, which can be changed to determine the importance of the indexes. Finally, via extensive experimentation, the efficiency and effectiveness of the proposed algorithm is verified. |
Year | DOI | Venue |
---|---|---|
2021 | 10.1007/978-3-030-96772-7_37 | PARALLEL AND DISTRIBUTED COMPUTING, APPLICATIONS AND TECHNOLOGIES, PDCAT 2021 |
Keywords | DocType | Volume |
Federated learning, Aggregation algorithm, Distributed learning | Conference | 13148 |
ISSN | Citations | PageRank |
0302-9743 | 0 | 0.34 |
References | Authors | |
0 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Zhenshan Bao | 1 | 0 | 2.03 |
Wei Bai | 2 | 0 | 0.34 |
Wenbo Zhang | 3 | 0 | 0.34 |