# REINFORCE

A Monte Carlo policy-gradient algorithm that increases the probability of actions in proportion to the return that followed them.

## What it is used for

Teaching and baseline use, and the core idea behind many policy-optimization methods.