Skip to yearly menu bar Skip to main content


Workshop

Extending Robust Adversarial Reinforcement Learning Considering Adaptation and Diversity

Hiroaki Shioya · Yusuke Iwasawa · Yutaka Matsuo

East Meeting Level 8 + 15 #4

Thu 3 May, 4:30 p.m. PDT

We propose two extensions to Robust Adversarial Reinforcement Learning. (Pinto et al., 2017) One is to add a penalty that brings the training domain closer to the test domain to the objective function of the adversarial agent. The other method trains multiple adversarial agents for one protagonist. We conducted experiments with the physical simulator benchmark task. The results show that our method improves performance in the test domain compared to the baseline.

Live content is unavailable. Log in and register to view live content