How to discretize continuous state-action spaces in Q-learning: A symbolic control approach.
Sadek Belamfedel AlaouiAdnane SaoudPublished in: CoRR (2024)
Keyphrases
- continuous state
- action space
- reinforcement learning
- control policies
- state space
- action selection
- continuous state spaces
- state action
- control problems
- optimal policy
- robot navigation
- markov decision processes
- policy search
- finite state
- real valued
- control policy
- optimal control
- state dependent
- function approximation
- learning algorithm
- continuous action
- single agent
- temporal difference
- cooperative
- motion control
- control system
- heuristic search
- reinforcement learning algorithms
- multi agent
- planning problems
- function approximators
- stochastic processes
- autonomous robots
- hidden state
- partially observable markov decision processes
- markov decision process
- model checking
- arrival rate
- search algorithm
- machine learning
- state transition