Build a Recycling Robot 

5 years ago .  Read Time: 1 min 30 sec.

Recycling Robot: A Catalyst for Understanding Markov Decision Processes 🤖🔄

Let's embark on a journey into the realm of Markov Decision Processes (MDPs) and the intricacies of the Expectation Maximization (EM) algorithm, with a real-world companion – the Recycling Robot. ♻️

About the Project 📚

Markov Decision Processes serve as our mathematical compass for understanding decision-making in situations where outcomes blend randomness and the decision maker's control. Dive into the project to explore the essence of MDPs, solving them through the Value Iteration algorithm, and a practical example featuring the Recycling Robot.

Recycling Robot: How It Works 🌐

The Recycling Robot is no ordinary automaton; it's a strategic decision-maker. In the project, we delve into the Recycling Robot's role as it encounters three conditions: tapping with the left hand, tapping with the right hand, and a control condition where it does nothing. The project aims to understand the neural activity associated with these tapping conditions using the General Linear Model (GLM) analysis.

Getting Started 🚀

To embark on this intellectual voyage, ensure you have Python 3.x installed, a grasp of Markov Decision Processes and Reinforcement Learning, and basic knowledge of the Expectation Maximization algorithm.

Usage 🛠️

Ready to explore? Follow these steps:

Run main.py to delve into MDPs and solve them using the Value Iteration algorithm.

Peek into recycling_robot.py for a real-world MDP example featuring the Recycling Robot.

Check out the project report for an in-depth understanding of the Expectation Maximization algorithm.

Contributing 🤝

Contribute to the brilliance of open source! Your suggestions and enhancements are the backbone of community growth. Follow these steps:

Fork the project.

Create a feature branch.

Commit your changes.

Push to the branch.

Open a pull request.

Don't forget to give the project a star – your support is invaluable! ⭐

References 📖

Dive deeper into the theoretical foundations with these references:

Sutton, R. S., & Barto, A. G. (Reinforcement Learning: An Introduction)

Abhinav (Masters Thesis: Molecular Algorithms and Schemes for their Implementation using DNA)

Li, L. (A Unifying Framework For Computational Reinforcement Learning Theory)

Muppirala Viswa Virinchi, Abhishek Behera, Manoj GopalKrishnan (A reaction network scheme implementing the EM algorithm, and more about the Recycling Robot)

Embark on this educational journey, explore the algorithms, and contribute to the thriving world of knowledge! 🌍✨ For code.