Learning and Policy Search in Stochastic Dynamical Systems with Bayesian Neural Networks.
Stefan DepewegJosé Miguel Hernández-LobatoFinale Doshi-VelezSteffen UdluftPublished in: ICLR (Poster) (2017)
Keyphrases
- dynamical systems
- policy search
- neural network
- predictive state representations
- learning algorithm
- policy search methods
- learning tasks
- learning problems
- partially observable
- partially observable markov decision processes
- state space
- reinforcement learning methods
- reinforcement learning
- hidden variables
- hidden state
- continuous state
- supervised learning