Proximal policy optimization (PPO)

My notes from learning to implement PPO including trust regions, importance sampling, and other topics.

Physics informed neural networks

Physics informed neural networks use additional loss terms based on the dynamics of the function being learned to improve performance for learning complex systems.