Weighted mesh algorithms for general Markov decision processes: Convergence and tractability.
Denis BelomestnyJohn SchoenmakersPublished in: CoRR (2024)
Keyphrases
- markov decision processes
- stochastic shortest path
- policy iteration
- computational complexity
- theoretical justification
- factored mdps
- reinforcement learning
- convergence rate
- learning algorithm
- state space
- finite state
- reachability analysis
- optimal policy
- dynamic programming
- search algorithm
- planning under uncertainty
- incremental algorithms
- continuous state spaces
- decision theoretic planning
- action sets
- transition matrices