Reinforcement Learning for Omega-Regular Specifications on Continuous-Time MDP.
Amin FalahShibashis GuhaAshutosh TrivediPublished in: ICAPS (2023)
Keyphrases
- reinforcement learning
- state space
- markov decision processes
- markov decision process
- optimal policy
- optimal control
- semi markov decision process
- reward function
- reinforcement learning algorithms
- action sets
- markov chain
- policy iteration
- function approximation
- dynamic programming
- average reward
- state and action spaces
- finite state
- stationary policies
- partially observable
- model free
- infinite horizon
- dynamical systems
- temporal difference
- delay insensitive
- high level
- markov decision problems
- heuristic search
- learning algorithm
- markov processes
- supervised learning
- reinforcement learning methods
- planning under uncertainty
- learning agent
- machine learning
- formal specification
- iterative learning control
- function approximators
- action space
- partial observability
- multi agent
- state abstraction
- decision theoretic planning
- learning problems