Introduction:

The project titled "Reinforcement Learning in Super Mario Bros" is dedicated to harnessing advanced artificial intelligence methodologies to create a self-sufficient agent proficient in playing the iconic video game, Super Mario Bros. Through the application of Reinforcement Learning (RL) techniques, the project endeavors to empower an AI agent to adeptly maneuver the game's intricate landscapes, surmount obstacles, and successfully navigate levels with growing competence. Employing RL algorithms, the agent will assimilate knowledge from its interactions, enabling it to make decisions geared towards maximizing rewards and enhancing its gameplay skills progressively.

Methodology - 

1. Initially, got familiarized with Python language and concepts of OOPS in Python.

2. Introduction to Machine learning(ML) and different Deep Learning Framework.

3. Familiarize with OpenAIGym and Machine learning models like Proximal Policy Optimization (PPO)

4. To apply the above concepts, make a model For Cartpole Game. 

5. After having sufficient knowledge build a mode for Super Mario Bros.

Real-life applications:

1. Autonomous Systems: RL is used in autonomous vehicles (self-driving cars, drones) and robots to learn how to navigate complex environments, make decisions, and avoid obstacles.

2. Game Playing: RL algorithms have been used to create game-playing agents that can excel in games like chess, Go, and video games. Notable examples include AlphaGo, which defeated human Go champions.

3. Finance and Trading: RL is used in algorithmic trading to optimize trading strategies, manage portfolios, and make investment decisions based on market conditions and historical data.

Problems faced:

To address unexpected errors in the code caused by incompatible library versions.

Mentors:

Contributors:   

TEAM 1

TEAM 2

TEAM 3

RL in Super Mario Bros Project Team