K-Means Clustering: An Explorable Explainer by Yi Zhe Ang

This interactive article presents a so-called “explorable explainer” of the k-means clustering algorithm. It attempts to push the envelope of how visual explanations can be designed on an interactive computational medium such as the web.

This article explains the algorithm using a “scrollytelling” piece. As the reader scrolls through the article, not only do the visuals animate and transition accordingly with the text, the reader is also prompted to interact with these geometric visualizations of the algorithm, supplementing their intuition through hands-on experimentation with its inner workings.

As the article progresses and introduces more concepts, it also slowly introduces corresponding knobs and dials that the reader can play around with. The article finally concludes with a dashboard or sandbox that comprises all the interactive controls that came before, giving the reader free rein to probe around and dissect the algorithm to improve their own understanding of its details as they please.

#