Leveraging Reward Gradients For Reinforcement Learning in Differentiable Physics Simulations.
Sean GillenKatie BylPublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- function approximation
- state space
- model free
- computer science
- total reward
- reward function
- markov decision processes
- reinforcement learning algorithms
- machine learning
- objective function
- partially observable environments
- reinforcement learning methods
- eligibility traces
- transfer learning
- multi agent
- learning algorithm
- temporal difference
- optimal control
- loss function
- learning capabilities
- partially observable markov decision processes
- physical models
- optimal policy
- robotic control
- dynamic programming
- action selection
- simulation environment
- numerical simulations
- temporal difference learning
- learning problems
- multi agent reinforcement learning
- qualitative physics
- reward shaping