Stochastic Policy

reinforcement-learning

Stochastic Policy

Given a state space $S$ and an action space $A$, a stochastic policy is a map

$$\pi :S\to \Delta (A),$$

where $\Delta (A)$ is the set of probability distributions over $A$. Equivalently,

$$\pi (a\mid s)=\mathrm{Pr}(A_t=a\mid S_t=s)$$

is the probability of choosing action $a\in A$ in state $s\in S$.