Empirical Value Iteration for approximate dynamic programming.
William B. HaskellRahul JainDileep M. KalathilPublished in: ACC (2014)
Keyphrases
- approximate dynamic programming
- policy iteration
- dynamic programming
- markov decision processes
- reinforcement learning
- linear program
- average cost
- stochastic dynamic programming
- optimal policy
- factored mdps
- state space
- model free
- least squares
- markov decision process
- control policy
- fixed point
- step size
- optimal control
- infinite horizon
- temporal difference
- linear programming
- heuristic search
- machine learning
- finite state
- multistage
- multiresolution
- computational complexity
- decision theoretic
- data mining
- average reward
- markov decision problems
- search space