What is reinforcement learning policy?

Question

I saw words like:

The policy determines the student’s current behavior. Roughly speaking, politics is a comparison of perceived environmental conditions with the actions that must be taken in these states.

But still not completely understood. What exactly is reinforcement learning policy?

Answer 1

The definition is correct, although not immediately obvious, if you see it for the first time. Let me say this: politics is an agent’s strategy.

, , , , (x, y), . :

• - .
• -

• - , :

  • , ( № 1)
  • - ( № 2).
  • "" ( № 3).

, , , . RL . ( , ):

.

(MDP) (S, A, P, R, y), :

• S -
• A -
• P - ( )
• R -
• y - 0 1

π , . , (, ). .

David Silver RL YouTube. .

Answer 2

, π - , s a. : π(s) → a

, , a , s.

, . a .

, RL , .

Answer 3

: - "" . , - s, a ? :

state----action----probability/'goodness' of taking the action
  1         1                     0.6
  1         2                     0.4
  2         1                     0.3
  2         2                     0.7

1, ( ) 1. 2, 2.