Coping with Bad Agent Interaction Protocols When Monitoring Partially Observable Multiagent Systems.
Davide AnconaAngelo FerrandoLuca FranceschiniViviana MascardiPublished in: PAAMS (2018)
Keyphrases
- multiagent systems
- partially observable
- agent interactions
- multi agent
- multi agent systems
- reinforcement learning
- state space
- markov decision processes
- decision problems
- dynamical systems
- partial observability
- intelligent agents
- agent societies
- autonomous agents
- infinite horizon
- interaction protocols
- markov decision problems
- software agents
- belief state
- cooperative
- coalition formation
- distributed learning
- decision model
- partially observable markov decision processes
- optimal policy
- dec pomdps
- complex domains
- orders of magnitude
- decentralized control
- decision making