Decentralized Planning in Stochastic Environments with Submodular Rewards.
Rajiv Ranjan KumarPradeep VarakanthamAkshat KumarPublished in: AAAI (2017)
Keyphrases
- fully observable
- planning problems
- stochastic domains
- decentralized decision making
- assembly systems
- bandit problems
- peer to peer
- uncertain environments
- reinforcement learning
- goal oriented
- real world
- dynamic environments
- markov decision processes
- cooperative
- high order
- greedy algorithm
- heuristic search
- ai planning
- planning process
- motion planning
- distributed systems
- open systems
- blocks world
- belief space
- partial observability
- monte carlo
- state space
- multi agent
- multi armed bandits
- long term and short term