Contrastive Initial State Buffer for Reinforcement Learning.
Nico MessikommerYunlong SongDavide ScaramuzzaPublished in: CoRR (2023)
Keyphrases
- initial state
- reinforcement learning
- optimal policy
- state space
- markov decision process
- situation calculus
- markov decision processes
- action theories
- function approximation
- stationary distribution
- partially observable markov decision processes
- decision problems
- goal state
- dynamic programming
- markov chain
- heuristic search
- probability distribution
- finite state
- long run
- multi agent
- average cost
- conformant planning
- sufficient conditions
- learning algorithm
- infinite horizon
- state variables
- optimal control
- belief space
- machine learning
- heuristic function
- function approximators
- domain specific