Home▸AI & Machine Learning▸DBSCAN: Finding Clusters of Any Shape by Density, Not Distance (2D)

DBSCAN: Finding Clusters of Any Shape by Density, Not Distance (2D)

A flat top-down 2D DBSCAN explorer: watch core, border and noise points get labeled live as you drag eps and min_samples, with a density-reachability graph and a k-distance plot for picking eps.

AI & Machine Learning2DModerate60 FPS📱 Mobile-adapted⇄ 3D version

This 2D companion runs the identical DBSCAN algorithm as the 3D version through a flat top-down canvas: pick a dataset shape, drag eps and min_samples, and watch every point get relabeled core, border or noise in real time, with an optional density-reachability graph and a k-distance plot for choosing eps.

⚙ Under the hood

2D DBSCAN explorer: core/border/noise classification recomputed live from eps and min_samples, a reachability graph of ε-connected core points, and a k-distance plot for picking eps.

dbscanclusteringunsupervised learningepsmin_samplesk-distance plot

2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install