Reinforcement Learning for Gaming - Full Python Course

Overview

Embark on a comprehensive 9-hour Python course focused on applying Reinforcement Learning (RL) to gaming. Learn best practices for training RL models using Python and Stable Baselines 3 across four popular games: Mario, Doom, Street Fighter, and Chrome Dino. Begin with setting up environments, preprocessing data, and building RL models for each game. Progress through advanced topics like OpenAI Gym integration, hyperparameter tuning, curriculum learning, and reward shaping. Gain hands-on experience in creating custom environments, fine-tuning models, and testing RL agents. By the end of this extensive tutorial, master the skills to apply machine learning techniques to various gaming scenarios, enhancing your understanding of AI in interactive environments.

Syllabus

- START
- MARIO
- Mario Mission 1 - Setup Mario
- Mario Mission 2 - Preprocess Environment
- Mario Mission 3 - Build the RL Model
- Mario Mission 4 - Run the RL Model Live
- DOOM
- Doom Mission 1 - Get Vizdoom Working
- Doom Mission 2 - Setup OpenAI Gym Environment
- Doom Mission 3 - Train the RL Agent
- Doom Mission 4 - Test the RL Agent
- Doom Mission 5 - Training for Other Levels
- Doom Mission 6 - Curriculum Learning and Reward Shaping
- STREETFIGHTER
- Streetfighter Mission 1 - Setup Streetfighter
- Streetfighter Mission 2 - Preprocessing
- Streetfighter Mission 3 - Hyperparameter Tuning
- Streetfighter Mission 4 - Fine Tune the Model
- Streetfighter Mission 5 - Testing the Model
- DINO
- Dino Mission 1 - Install and Setup Dependencies
- Dino Mission 2 - Create a Custom OpenAI Gym Environment
- Dino Mission 3 - Train the RL Model
- Dino Mission 4 - Get the Model to Smash Chrome Dino
- Wrap Up