It's a wrap! Today was our final presentation, and I managed to finish the largest part of my final project. I made a poster-style dashboard about how the Machine Learning Algorythm Random Forest work, with the idea of turning it into a small series for other ML algorithms later on. The dashboard itself is split into two parts: the top section is designed to be creative and exploratory, while the lower section provides the detailed technical information.
Right now, the model trains 9 trees, but I only visualized three of them to keep the layout clean. In the top section, you can pick a mushroom and see how the forest makes its decisions. If you hover over the nodes and leaves, it shows the specific question the tree is asking at that point, which makes the process quite transparent. Further down, the individual steps are explained with a more technical approach. I'm hoping this mix is both interesting and easy to learn from.
I’m happy with how it looks so far, but there are still a few things to fix. For example, I want users to be able to select the mushrooms by clicking directly on them instead of using a dropdown filter, to make it more hands-on. I also need to make sure the user journey feels intuitive.
If you want to check it out, you can find the dashboard here: Random Forest. I'd be really happy to get some feedback!
