Automating Discovery of Physics-Informed Neural State Space Models via Learning and Evolution.
Elliott SkomskiJan DrgonaAaron TuorPublished in: L4DC (2021)
Keyphrases
- reinforcement learning
- prior knowledge
- state space
- learning algorithm
- accurate models
- learning process
- probabilistic model
- search space
- neural network
- learning models
- stochastic domains
- recurrent networks
- artificial intelligence
- learned models
- learning tasks
- mathematical models
- neural computation
- physical models
- neural nets
- experimental data
- heuristic search
- complex systems
- model selection
- active learning
- computer science