BR-DeFedRL: Byzantine-Robust Decentralized Federated Reinforcement Learning with Fast Convergence and Communication Efficiency.
Jing QiaoZuyuan ZhangSheng YueYuan YuanZhipeng CaiXiao ZhangJu RenDongxiao YuPublished in: INFOCOM (2024)
Keyphrases
- reinforcement learning
- fault tolerant
- distributed systems
- multi agent
- decentralized control
- learning process
- state space
- communication overhead
- cooperative
- learning algorithm
- stochastic approximation
- reinforcement learning algorithms
- highly efficient
- function approximation
- convergence rate
- peer to peer
- digital libraries
- learning problems
- markov decision processes
- information sharing
- wireless networks
- model free
- communication protocol
- dynamic programming
- wireless sensor networks