Vehicular Cooperative Perception Through Action Branching and Federated Reinforcement Learning.
Mohamed K. Abdel-AzizCristina PerfectoSumudu SamarakoonMehdi BennisWalid SaadPublished in: IEEE Trans. Commun. (2022)
Keyphrases
- cooperative
- reinforcement learning
- action selection
- sensory inputs
- multi agent
- joint action
- partially observable domains
- multiagent reinforcement learning
- action space
- multi agent reinforcement learning
- function approximation
- reward shaping
- state action
- active exploration
- multi agent systems
- temporal difference
- digital libraries
- data sources
- reinforcement learning algorithms
- sensory data
- visual perception
- human actions
- continuous state
- transition model
- agent learns
- fitted q iteration
- partially observable
- heterogeneous networks
- model free
- urban areas
- human perception
- search tree
- branch and bound
- markov decision processes
- traffic congestion
- dynamic programming
- biologically inspired
- sensing actions
- game theory
- transfer learning
- optimal policy
- state space