Safe Reinforcement Learning with Linear Function Approximation

                 Sanae Amani 1 Christos Thrampoulidis 2 Lin F. Yang 1

             Abstract                action may lead to catastrophic results. Thus, safety in RL
                                 has become a serious issue that restricts the applicability of
    Safety in reinforcement learning has become in-      RL algorithms to many real-world systems. For example,
    creasingly important in recent years. Yet, exist-     in ...