Bayesian decomposition of multi-modal dynamical systems for reinforcement learning.
Markus KaiserClemens OtteThomas A. RunklerCarl Henrik EkPublished in: Neurocomputing (2020)
Keyphrases
- multi modal
- dynamical systems
- reinforcement learning
- state space
- reinforcement learning methods
- partially observable
- differential equations
- dynamic systems
- model free
- function approximation
- nonlinear dynamical systems
- multi modality
- bayesian networks
- reinforcement learning algorithms
- phase space
- high dimensional
- qualitative simulation
- agent environment
- audio visual
- video search
- optimal policy
- cross modal
- multi agent
- policy iteration
- partially observable markov decision processes
- temporal difference
- predictive state representations
- uni modal
- optimal solution
- multiple modalities
- monte carlo
- humanoid robot
- optimal control
- markov decision processes