Compact Policies for Fully-Observable Non-Deterministic Planning as SAT.
Tomas GeffnerHector GeffnerPublished in: CoRR (2018)
Keyphrases
- fully observable
- planning problems
- classical planning
- partially observable markov decision processes
- markov decision problems
- decentralized control
- partial observability
- domain independent
- state space
- heuristic search
- partially observable
- belief space
- optimal policy
- belief state
- planning under uncertainty
- planning domains
- markov decision process
- reinforcement learning
- finite state
- linear programming
- goal oriented
- decision problems
- markov decision processes
- deterministic domains
- hidden state
- decision theoretic
- dynamic programming
- dynamical systems
- multiagent systems
- ai planning
- expected utility
- sat solvers
- collective behavior
- search space
- search algorithm
- decision makers
- machine learning
- multi agent
- goal state
- reward function
- initial state
- utility function
- queueing networks