Policy Iteration

Concept

An alternative algorithm for solving infinite horizon MDPs that iteratively improves a policy by first evaluating the current policy and then improving it.

Mentioned in 1 video

Videos Mentioning Policy Iteration

Stanford AA203 Optimal and Learning-Based Control | Spring 2026 | Lecture 10: Reachibility Analysis

Stanford AA203 Optimal and Learning-Based Control | Spring 2026 | Lecture 10: Reachibility Analysis

Stanford Online

An alternative algorithm for solving infinite horizon MDPs that iteratively improves a policy by first evaluating the current policy and then improving it.