Accelerating Distributed Deep Reinforcement Learning by In-Network Experience Sampling.
Masaki FurukawaHiroki MatsutaniPublished in: PDP (2022)
Keyphrases
- reinforcement learning
- peer to peer
- distributed network
- computer networks
- communication overhead
- multi agent
- distributed systems
- communication cost
- network structure
- camera network
- peer to peer networks
- network nodes
- network traffic
- central server
- distributed environment
- network model
- sample size
- optimal policy
- mobile sensor
- distributed control
- random sampling
- concurrent processes
- reinforcement learning algorithms
- heterogeneous networks
- network architecture
- learning algorithm
- monte carlo
- lightweight
- web services