Reinforcement Learning Classes

Welcome to the Reinforcement Learning Classes repository. This repository contains a series of projects completed as part of the Reinforcement Learning course at Insper. Each project is designed to implement and evaluate various reinforcement learning algorithms across diverse environments, providing both theoretical insights and practical applications.

Contents

1. 01-taxi-driver-drone-drivers: Implementation of a taxi driver problem using the A* search algorithm.
2. 02-tic-tac-toe-jogo-do-galo: Tic-Tac-Toe game implemented with the Minimax algorithm.
3. 03-q-learning-implemented: Basic implementation of the Q-Learning algorithm applied to the Taxi Driver problem.
4. 04-q-learning-sarsa: Comparison between Q-Learning and SARSA algorithms.
5. 05-evaluate: Techniques for evaluating reinforcement learning models.
6. 06-stochastic-environment-frozen-lake: Application of Q-Learning and SARSA in the stochastic Frozen Lake environment.
7. 07-lunar-lander-droneiros-de-cabreuva: Solving the Lunar Lander problem using reinforcement learning and comparing Deep Q-Learning with DQN (Deep Q-Networks).
8. 08-compare: Comparative analysis of various reinforcement learning algorithms across different environments.
9. 09-intermediate-project-ornithopter: Intermediate project utilizing DQN and Double DQN across three different environments.
10. 10-final-project-droneiros-de-ibitinga: Final project featuring reinforcement learning research and hypotheses in an advanced drone search and rescue (SAR) mission simulation within a custom environment.