A Bayesian technique for task localization in multiple goal Markov decision processes.
James L. CarrollKevin D. SeppiPublished in: ICMLA (2004)
Keyphrases
- markov decision processes
- finite state
- state space
- transition matrices
- optimal policy
- dynamic programming
- factored mdps
- reinforcement learning
- finite horizon
- policy iteration
- markov decision process
- reinforcement learning algorithms
- decision processes
- average reward
- decision theoretic planning
- reachability analysis
- average cost
- partially observable
- infinite horizon
- model based reinforcement learning
- risk sensitive
- continuous state spaces
- action sets
- semi markov decision processes
- markov chain
- real time dynamic programming