Minimax PAC bounds on the sample complexity of reinforcement learning with a generative model.
Mohammad Gheshlaghi AzarRémi MunosHilbert J. KappenPublished in: Mach. Learn. (2013)
Keyphrases
- generative model
- reinforcement learning
- upper bound
- pac bayesian
- vc dimension
- worst case
- mistake bound
- probabilistic model
- expected error
- lower bound
- bayesian framework
- semi supervised
- prior knowledge
- sample complexity
- function approximation
- discriminative learning
- topic models
- posterior probability
- em algorithm
- temporal difference
- state space
- learning problems
- latent dirichlet allocation
- learning algorithm
- sample size
- pac learning
- markov decision processes
- expectation maximization
- transfer learning
- learning process
- discriminative models
- generative and discriminative models
- reward function
- markov chain monte carlo
- training set
- image processing