Enabling Efficient, Reliable Real-World Reinforcement Learning with Approximate Physics-Based Models.
Tyler WestenbroekJacob LevyDavid Fridovich-KeilPublished in: CoRL (2023)
Keyphrases
- real world
- reinforcement learning
- synthetic data
- data sets
- experimental data
- complex systems
- classification models
- function approximation
- temporal difference
- data structure
- wide range
- machine learning
- markov models
- parametric models
- database
- statistical models
- cost effective
- accurate models
- statistical model
- model selection
- supervised learning
- probability distribution
- probabilistic model
- dynamic programming
- prior knowledge
- learning process
- multi agent systems