Show and Tell - Making Interactive Machine Learning Explorers using Dash + Scikit-learn

Hey all! I’ve recently completed two apps that lets you explore different parameters for very popular learning algorithms, notably Least-Square regression and SVM Classifiers, through an intuitive UI. They are fully built in Dash (Components, Plotly Graphs, Server) and Scikit-Learn (Datasets, algorithms), limiting the dependencies to what you can easily find in PyPi.

I’d be happy if you give me some feedback on your thoughts, experience on using them, suggestions for future additions, etc. Since it’s an open-source project, you are welcome to fork it and add more features (more parameters to control, more datasets, or maybe even other types of algorithms?). PRs are welcome, and if you wish to make a completely new app let me know!

Dash SVM Explorer

Link to Repo | Demo App

Dash Regression Explorer

Link to Repo | Demo App


SVM app:

Regression app:


Brilliant App! I’ve been thinking of creating something very similar to allow our engineers to experiment with different ML techniques as many of them aren’t familiar with Python. What’s the largest dataset you’ve used with this so far?

Since they are demo apps I’ve mostly used smaller datasets (e.g. 500 records). However, this could easily work with very big datasets since Scikit-Learn has fairly optimized algorithms for statistical learning, and Plotly can render up to a million data points smoothly using WebGL.

Feel free to take a look at the source code and create an issue if something is unclear!