Reinforcement learning (RL) has had many successes in different tasks, but in practice, it often requires significant amounts of data or training time to learn high-performing policies. For complicated tasks, the learning ...