Federated Learning Incentive Mechanism Design via Shapley Value and Pareto Optimality
Federated learning (FL) is a distributed machine learning framework that can effectively help multiple players to use data to train federated models while complying with their privacy, data security, and government regulations. Due to federated model training, an accurate model should be trained, an...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-06-01
|
| Series: | Axioms |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-1680/12/7/636 |
| _version_ | 1797590282555883520 |
|---|---|
| author | Xun Yang Shuwen Xiang Changgen Peng Weijie Tan Zhen Li Ningbo Wu Yan Zhou |
| author_facet | Xun Yang Shuwen Xiang Changgen Peng Weijie Tan Zhen Li Ningbo Wu Yan Zhou |
| author_sort | Xun Yang |
| collection | DOAJ |
| description | Federated learning (FL) is a distributed machine learning framework that can effectively help multiple players to use data to train federated models while complying with their privacy, data security, and government regulations. Due to federated model training, an accurate model should be trained, and all federated players should actively participate. Therefore, it is crucial to design an incentive mechanism; however, there is a conflict between fairness and Pareto efficiency in the incentive mechanism. In this paper, we propose an incentive mechanism via the combination of the Shapley value and Pareto efficiency optimization, in which a third party is introduced to supervise the federated payoff allocation. If the payoff can reach Pareto optimality, the federated payoff is allocated by the Shapley value method; otherwise, the relevant federated players are punished. Numerical and simulation experiments show that the mechanism can achieve fair payoff allocation and Pareto optimality payoff allocation. The Nash equilibrium of this mechanism is formed when Pareto optimality payoff allocation is achieved. |
| first_indexed | 2024-03-11T01:18:21Z |
| format | Article |
| id | doaj.art-2c14e857982a41dabc7d7e02126b151d |
| institution | Directory Open Access Journal |
| issn | 2075-1680 |
| language | English |
| last_indexed | 2024-03-11T01:18:21Z |
| publishDate | 2023-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Axioms |
| spelling | doaj.art-2c14e857982a41dabc7d7e02126b151d2023-11-18T18:17:19ZengMDPI AGAxioms2075-16802023-06-0112763610.3390/axioms12070636Federated Learning Incentive Mechanism Design via Shapley Value and Pareto OptimalityXun Yang0Shuwen Xiang1Changgen Peng2Weijie Tan3Zhen Li4Ningbo Wu5Yan Zhou6School of Mathematics and Statistics, Guizhou University, Guiyang 550025, ChinaSchool of Mathematics and Statistics, Guizhou University, Guiyang 550025, ChinaState Key Laboratory of Public Big Data, College of Computer Science and Technology, Guizhou University, Guiyang 550025, ChinaState Key Laboratory of Public Big Data, College of Computer Science and Technology, Guizhou University, Guiyang 550025, ChinaCollege of Big Data and Information Engineering, Guizhou University, Guiyang 550025, ChinaSchool of Information, Guizhou University of Finance and Economics, Guiyang 550025, ChinaState Key Laboratory of Public Big Data, College of Computer Science and Technology, Guizhou University, Guiyang 550025, ChinaFederated learning (FL) is a distributed machine learning framework that can effectively help multiple players to use data to train federated models while complying with their privacy, data security, and government regulations. Due to federated model training, an accurate model should be trained, and all federated players should actively participate. Therefore, it is crucial to design an incentive mechanism; however, there is a conflict between fairness and Pareto efficiency in the incentive mechanism. In this paper, we propose an incentive mechanism via the combination of the Shapley value and Pareto efficiency optimization, in which a third party is introduced to supervise the federated payoff allocation. If the payoff can reach Pareto optimality, the federated payoff is allocated by the Shapley value method; otherwise, the relevant federated players are punished. Numerical and simulation experiments show that the mechanism can achieve fair payoff allocation and Pareto optimality payoff allocation. The Nash equilibrium of this mechanism is formed when Pareto optimality payoff allocation is achieved.https://www.mdpi.com/2075-1680/12/7/636federated learningShapley valuePareto optimalityNash equilibrium |
| spellingShingle | Xun Yang Shuwen Xiang Changgen Peng Weijie Tan Zhen Li Ningbo Wu Yan Zhou Federated Learning Incentive Mechanism Design via Shapley Value and Pareto Optimality Axioms federated learning Shapley value Pareto optimality Nash equilibrium |
| title | Federated Learning Incentive Mechanism Design via Shapley Value and Pareto Optimality |
| title_full | Federated Learning Incentive Mechanism Design via Shapley Value and Pareto Optimality |
| title_fullStr | Federated Learning Incentive Mechanism Design via Shapley Value and Pareto Optimality |
| title_full_unstemmed | Federated Learning Incentive Mechanism Design via Shapley Value and Pareto Optimality |
| title_short | Federated Learning Incentive Mechanism Design via Shapley Value and Pareto Optimality |
| title_sort | federated learning incentive mechanism design via shapley value and pareto optimality |
| topic | federated learning Shapley value Pareto optimality Nash equilibrium |
| url | https://www.mdpi.com/2075-1680/12/7/636 |
| work_keys_str_mv | AT xunyang federatedlearningincentivemechanismdesignviashapleyvalueandparetooptimality AT shuwenxiang federatedlearningincentivemechanismdesignviashapleyvalueandparetooptimality AT changgenpeng federatedlearningincentivemechanismdesignviashapleyvalueandparetooptimality AT weijietan federatedlearningincentivemechanismdesignviashapleyvalueandparetooptimality AT zhenli federatedlearningincentivemechanismdesignviashapleyvalueandparetooptimality AT ningbowu federatedlearningincentivemechanismdesignviashapleyvalueandparetooptimality AT yanzhou federatedlearningincentivemechanismdesignviashapleyvalueandparetooptimality |