CityManager Reinforcement Learning is an interactive web-based game that combines the fun of city-building with the advanced technology of artificial intelligence. 🌆👾 Build your city, deploy AI agents, and watch them learn and evolve in real-time!
-
For more details, checkout our Short Report!
-
You can also chekcout our Teaser Video for a better understanding of the project.
-
You can even have access to our PowerPoint Presentation for a quick demo.
- Interactive City Building: Place houses, offices, and restaurants to create your city.
- AI Agents: Utilize Q-Learning (red agents) and Deep Q-Network (DQN - blue agents) algorithms.
- Simulation Control: Adjust the number of agents, exploration rate decay, and episode length.
- Dynamic Weather: Influence the game by changing the weather conditions.
- Real-Time Learning Visualization: Monitor the agents' learning progress through live graphs.
- Built with React, Python, and Java: A harmonious blend of technologies for a seamless experience.
- Dockerized Deployment & CI/CD Pipeline: Ensuring consistent performance and ease of updates.
- Visit the Game: Head over to CityManager Game to start playing. Build Your City: Strategically place buildings and set up your city. Deploy AI Agents: Choose your agents and watch them navigate your city. Customize & Observe: Change game settings and observe how agents adapt and learn.
This application is containerized using Docker and uses GitHub Workflows for continuous integration and deployment, ensuring a reliable and consistent experience. Feedback and Contributions 💡
https://trello.com/w/citymanagergr5
