Policy gradient bayesian robust optimization for imitation learning

摘要: The difficulty in specifying rewards for many real-world problems has led to an increased focus on learning rewards from human feedback, such as demonstrations. However, there …