Exploiting Causal Graph Priors with Posterior Sampling for Reinforcement Learning.
Mirco MuttiRiccardo De SantiMarcello RestelliAlexander MarxGiorgia RamponiPublished in: CoRR (2023)
Keyphrases
- causal graph
- reinforcement learning
- bayesian framework
- metropolis hastings algorithm
- planning problems
- proposal distribution
- markov chain monte carlo
- causal models
- state space
- posterior distribution
- particle filter
- monte carlo
- posterior probability
- learning algorithm
- state variables
- maximum a posteriori
- markov decision processes
- heuristic search
- probability distribution
- decision problems
- plan generation
- multi agent
- learning problems
- optimal control
- gaussian process
- probabilistic model
- machine learning