Sample-Efficient Reinforcement Learning for Linearly-Parameterized MDPs with a Generative Model.
Bingyan WangYuling YanJianqing FanPublished in: CoRR (2021)
Keyphrases
- generative model
- reinforcement learning
- markov decision processes
- probabilistic model
- em algorithm
- bayesian framework
- optimal policy
- reward function
- reinforcement learning algorithms
- prior knowledge
- dynamic programming
- action space
- fully bayesian
- semi supervised
- probability distribution
- latent dirichlet allocation
- markov chain monte carlo
- image processing
- state and action spaces
- machine learning
- discriminative models
- generative and discriminative models
- partially observable
- function approximation
- topic models
- learning algorithm