A simple and intuitive explanation of DBSCAN (Density-Based Spatial Clustering of Applications with Noise), a clustering algorithm that can identify outliers, extract new features, compress data, and perform novelty detection. The article provides a fast implementation of DBSCAN in Python.