Here’s a quick update on my traffic prediction tool for Wolfsburg. This project employs machine learning techniques to better understand and predict traffic patterns in the city.

You can see it in action here: http://www.corrix.net

The heart of this project is my unsupervised learning algorithm. "Unsupervised" means it learns independently, without needing pre-processed training data. While capable of various tasks, such as playing video games, I've directed its focus to a practical application: analyzing Wolfsburg's traffic data. This evolving system continuously adapts, observing and learning from the city's traffic patterns to enhance my daily commute.

Thanks to the system's efficiency, I can offer it for free in Wolfsburg. It runs on a modest cloud setup, keeping costs super low.

Currently, the algorithm focuses on major roads, learning from their daily traffic patterns and any irregularities. I've prioritized establishing reliable data collection and fine-tuning the algorithm's response to this data. It's designed to predict traffic jams only when it has high confidence in its analysis and will adapt and change to implement its increasing knowledge.

I'm steadily adding new roads and routes to improve coverage. Behind the scenes, the algorithm is also being refined for better performance.

I invite everyone to try it out! It’s free, easy to use and who knows, it might make your commute a little smoother. Any feedback? Feel free to reach out to me!