In reinforcement learning, If the range of observation is ±inf, how to normalize the observation?

I want to improve the performance of the agent. I think the critical point of improving the agent is to normalize the observation and action. However, I don't know how to normalize them when the...

12 February 2022 2,088 9 View