Symbolic Opportunistic Policy Iteration for Factored-Action MDPs.
Aswin RaghavanRoni KhardonAlan FernPrasad TadepalliPublished in: NIPS (2013)
Keyphrases
- policy iteration
- markov decision processes
- discounted reward
- state space
- action space
- optimal policy
- factored mdps
- markov decision problems
- reinforcement learning
- factored markov decision processes
- algebraic decision diagrams
- average reward
- temporal difference
- markov decision process
- model free
- sample path
- fixed point
- least squares
- finite state
- infinite horizon
- policy evaluation
- approximate dynamic programming
- initial state
- action selection
- finite horizon
- partially observable
- reinforcement learning algorithms
- state and action spaces
- transition matrices
- dynamic programming
- decision processes
- average cost
- basis functions
- linear programming
- stochastic games
- convergence rate
- decision problems
- markov chain
- multi agent