On the Hidden Biases of Policy Mirror Ascent in Continuous Action Spaces.
Amrit Singh BediSouradip ChakrabortyAnjaly ParayilBrian M. SadlerPratap TokekarAlec KoppelPublished in: CoRR (2022)
Keyphrases
- action space
- state space
- markov decision processes
- state and action spaces
- real valued
- reinforcement learning
- control policies
- continuous state spaces
- continuous state
- action selection
- state action
- stochastic processes
- markov decision process
- optimal policy
- markov decision problems
- reinforcement learning problems
- single agent
- dynamic programming
- learning algorithm
- skill learning
- continuous action
- function approximators
- reinforcement learning algorithms
- finite state
- state variables
- markov chain
- probabilistic model
- mobile robot