3 ms·

Now you are asking for a perfect modeling of the system. Reinforcement learning works by discovering boundaries.