I built RandomForestGenerator, a single-page tool to train a Random Forest on a CSV entirely in the browser, inspect results/diagnostics, and export a portable artifact you can reuse elsewhere.
Repo: https://github.com/jonaraphael/RandomForestGenerator
I've uploaded the titanic dataset for you to try as a quick demo, or drop in your own CSV.
I used Jeremy Howard's guidance on tabular ML to try to make it as useful as possible and provide best practices baked in for you.
I often want a quick “drop a CSV → get a baseline + sanity-check signals → export something reusable” loop, without needing to spin up a Python environment. So now that I've built this for myself, I find that even my non-ML colleagues are excited to try it on their own spreadsheets!
Everything happens locally in the browser tab--you could even disconnect from Wi-Fi after loading the HTML itself!
Just a head's up that large datasets can be slow/heavy in-browser (there’s a maxRows subsample knob) to try to mitigate.
Feedback I’d love: • Is this useful? • How could it be better? • A star if you think you might use it someday!
Thank you for checking out my little project!!! Have a great weekend