Transcription of An Introduction to Markov Decision Processes
{{id}} {{{paragraph}}}
MDP Tutorial - 1An Introduction toMarkov Decision ProcessesBob GivanRon ParrPurdue UniversityDuke UniversityMDP Tutorial - 2 OutlineMarkov Decision Processes defined (Bob) Objective functions PoliciesFinding Optimal Solutions (Ron) Dynamic programming Linear programmingRefinements to the basic model (Bob) Partial observability Factored representationsMDP Tutorial - 3 Stochastic Automata with UtilitiesAMarkov Decision Process (MDP) modelcontains: A set of possible world statesS A set of possible actionsA A real valued reward functionR(s,a) A descriptionTof each action s effects in each :the effects of an actiontaken in a state depend only on that state and not on theprior Tutorial - 4 Stochastic Automata with UtilitiesAMarkov Decision Process (MDP) modelcontains: A set of possible world statesS A set of possible actionsA A real valued reward functionR(s,a) A descriptionTof each action s effects in each :the effects of an actiontaken in a state depend only on that state and not on theprior Tutorial - 5 Representing ActionsDeterministic Actions: T : For each state and action we specify a new Actions: T:For each state and a
• models preference for shorter solutions ... In AI problems, the “state space” is typically • astronomically large • described implicitly, not enumerated • decomposed into factors, or aspects of state Issues raised: • How can we represent reward and action behaviors
Domain:
Source:
Link to this page:
Please notify us if you found a problem with this document:
{{id}} {{{paragraph}}}