C3_W3 Quiz (State-action value function) Question 2

[shinli256]

The question reads:

You are controlling a robot that has 3 actions: ← (left), → (right) and STOP. From a given state s, you have computed Q(s, ← ) = -10, Q(s, → ) =-20, Q(s, STOP ) = 0.

What is the optimal action to take in state s?

I see what it’s asking and have no problem with the answer. However, I do think we can specify what “optimal” means here, more specifically, whether we are optimizing the reward of one action or the total return.

[rmwkwok]

Hello Joe,

Would you think differently after reading this again, that Q is defined to be covering “after that” too?

Cheers,
Raymond

[shinli256]

Thanks providing the material.

I suppose they’re consistent in using the actually useful explanation for “optimal”.