Robot Playing Kendama with Model-Based and Model-Free Reinforcement Learning

Shidi Li

Several model-based and model-free methods have been proposed for the robot trajectory learning task. Both approaches have their benefits and drawbacks. They can usually complement…