Reinforcement Learning in Factored Action Spaces using Tensor Decompositions.
Anuj MahajanMikayel SamvelyanLei MaoViktor MakoviychukAnimesh GargJean KossaifiShimon WhitesonYuke ZhuAnimashree AnandkumarPublished in: CoRR (2021)
Keyphrases
- action space
- state space
- reinforcement learning
- state and action spaces
- continuous state
- markov decision processes
- real valued
- control policies
- dynamic programming
- reinforcement learning algorithms
- reinforcement learning methods
- state action
- continuous state spaces
- markov decision process
- function approximators
- markov chain
- action selection
- heuristic search
- state variables
- model free
- skill learning
- optimal policy
- higher order
- particle filter
- markov decision problems
- multi agent
- policy iteration
- function approximation
- dynamical systems
- single agent
- belief state
- finite state
- optimal control
- search space
- cooperative