Extrinsicaly Rewarded Soft Q Imitation Learning with Discriminator

no code implementations • 30 Jan 2024 • Ryoma Furuyama, Daiki Kuyoshi, Satoshi Yamane

In order to make this algorithm more robust to distribution shift, we propose more efficient and robust algorithm by adding to this method a reward function based on adversarial inverse reinforcement learning that rewards the agent for performing actions in status similar to the demo.

Imitation Learning Q-Learning +1