A classical approach to any reinforcement learning (RL) problem is to explore and to exploit. Explore the most rewarding way that reaches the target and keep on exploiting a certain action; exploration is hard. Without proper reward functions, the algorithms can end up chasing their own tails to eternity. When we say rewards, think of them as mathematical functions crafted carefully to nudge the algorithm. To be more precise, consider teaching a robotic arm or an AI playing a strategic game like Go or Chess to reach a target on its own.
Read more: https://analyticsindiamag.com/exploration-reinforcement-learning/