The computational complexity of avoiding spurious states in state space abstraction.
Sandra ZillesRobert C. HoltePublished in: Artif. Intell. (2010)
Keyphrases
- state space
- state abstraction
- computational complexity
- state variables
- state transition
- initial state
- goal state
- reinforcement learning
- belief state
- markov decision processes
- heuristic search
- markov decision problems
- decision theoretic planning
- reinforcement learning algorithms
- finite state automaton
- optimal policy
- special case
- dynamic programming
- markov chain
- np complete
- continuous state spaces
- state information
- low complexity
- decision problems
- computational efficiency
- particle filter
- high computational complexity
- planning problems
- macro actions
- partially observable
- state transitions
- stochastic domains
- action space
- heuristic function
- reward function
- monte carlo
- rate distortion
- computationally efficient
- np hard
- high level
- genetic algorithm
- continuous time markov process
- storage requirements
- function approximation
- computational cost