Journal
IEEE TRANSACTIONS ON CYBERNETICS
Volume 50, Issue 10, Pages 4214-4227Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TCYB.2019.2906574
Keywords
Privacy; Differential privacy; Base stations; Reinforcement learning; Multi-agent systems; Buildings; Routing; Agent advising; differential privacy; multiagent reinforcement learning (MARL)
Categories
Funding
- Australian Research Council Linkage Project [LP170100123]
- NSF [IIS-1526499, IIS-1763325, CNS-1626432]
- NSFC [61672313]
- Australian Research Council [LP170100123] Funding Source: Australian Research Council
Ask authors/readers for more resources
Agent advising is one of the key approaches to improve agent learning performance by enabling agents to ask for advice between each other. Existing agent advising approaches have two limitations. The first limitation is that all the agents in a system are assumed to be friendly and cooperative. However, in the real world, malicious agents may exist and provide false advice to hinder the learning performance of other agents. The second limitation is that the analysis of communication overhead in these approaches is either overlooked or simplified. However, in communication-constrained environments, communication overhead has to be carefully considered. To overcome the two limitations, this paper proposes a novel differentially private agent advising approach. Our approach employs the Laplace mechanism to add noise on the rewards used by student agents to select teacher agents. By using the differential privacy technique, the proposed approach can reduce the impact of malicious agents without identifying them. Also, by adopting the privacy budget concept, the proposed approach can naturally control communication overhead. The experimental results demonstrate the effectiveness of the proposed approach.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available