Author: Oualid Missaoui

This repository showcases a comprehensive implementation of exact dynamic programming algorithms designed to efficiently solve Markov Decision Processes (MDPs). Through this implementation, users can explore and understand the inner workings of policy evaluation, policy iteration, and value iteration algorithms, pivotal techniques used in reinforcement learning and decision-making tasks.

Moreover, this repository goes beyond mere theoretical explanations by providing practical demonstrations. Specifically, it demonstrates how these algorithms operate in practice by applying them to the challenging and popular Frozen Lake environment from the OpenAI Gym framework. This real-world illustration enables users to witness the algorithms in action and gain a hands-on understanding of their behavior and efficacy.

By offering a combination of theoretical explanations, robust implementations, and concrete examples, this repository aims to serve as a valuable resource for anyone looking to dive into the realm of dynamic programming, reinforcement learning, and solving complex decision-making problems.