Infinite horizon average cost dynamic programming subject to ambiguity on conditional distribution.
Ioannis TzortzisCharalambos D. CharalambousThemistoklis CharalambousPublished in: CDC (2015)
Keyphrases
- infinite horizon
- average cost
- dynamic programming
- conditional distribution
- random variables
- optimal policy
- optimal control
- finite horizon
- gaussian process
- long run
- markov decision processes
- probability distribution
- joint distribution
- single item
- production planning
- lead time
- multistage
- state space
- marginal distributions
- dec pomdps
- partially observable
- markov decision process
- initial state
- holding cost
- fixed cost
- inventory models
- linear programming
- policy iteration
- latent variable models
- finite state
- conditional probabilities
- latent variables
- reinforcement learning
- markov decision problems
- machine learning
- decision making
- stationary policies
- dirichlet process
- graphical models
- posterior distribution