Clustering with DBSCAN
August 5, 2024
Clustering in 2D is an interesting task: Given a set of points, how can we find groups (or clusters) that make sense?
One of several approaches is DBSCAN (Density-based spatial clustering of applications with noise).
Based on a given distance ε (which defines if a point is close enough to another), the algorithm looks for "dense" groups of points with a minimum amount of points.
You can try it yourself here:
You can find the source code on Github: dbscan-p5
Happy clustering!
