Notes Clustering by fast search and find of density peaks

We propose an approach based on the idea that cluster centers are characterized
By a higher density than their neighbors and by a relatively large distance from points
Higher densities. This idea forms the basis of a clustering procedure in which the number
Clusters arises intuitively, outliers are automatically spotted and excluded from the analysis, and
Clusters are recognized regardless of their shape and of the dimensionality of the space in which
They are embedded.

The idea source is direct: 1. The center density of the cluster is high; 2. The distance between the centers of the high density should be relatively far. Exception points are excluded and irrelevant to the shape.

The question is, how is density defined?

Dc is the phase distance and threshold. The number of points in this range. The selection of this value is robust.

In short, find the distance between the closest vertex and the closest vertex that is higher than the I node density.

Therefore, if the density is large, and the distance from other points with a higher density is also far away, it is probably the center of a cluster.

The point with the highest density is defined as the distance from its farthest point. By default, it is the center of a cluster.

We will continue to add this item tomorrow.

 

Notes Clustering by fast search and find of density peaks