Practical Probabilistic Model-based Deep Reinforcement Learning by Integrating Dropout Uncertainty and Trajectory Sampling.
Wenjun HuangYunduan CuiHuiyun LiXinyu WuPublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- uncertain data
- model free
- probability theory
- conditional probabilities
- function approximation
- partial observability
- data driven
- dynamic programming
- data sets
- temporal difference
- markov decision processes
- sample size
- bayesian networks
- probability measures
- robotic control
- decision theory
- random sampling
- belief networks
- monte carlo
- optimal policy
- learning classifier systems
- transfer learning
- sampling methods
- probabilistic logic
- inference process
- handling uncertainty
- real world