Efficient Off-Policy Meta-Reinforcement Learning via Probabilistic Context Variables.
Kate RakellyAurick ZhouChelsea FinnSergey LevineDeirdre QuillenPublished in: ICML (2019)
Keyphrases
- reinforcement learning
- probabilistic model
- generative model
- computationally efficient
- machine learning
- information retrieval
- function approximation
- cost effective
- continuous valued
- database
- temporal context
- uncertain data
- context sensitive
- conditional probabilities
- graphical models
- state space
- hidden markov models
- artificial neural networks
- high dimensional
- data streams