Core point

As explained in Data Without Labels

A core point is a DBSCAN observation having at least minPts points, including itself, within epsilon distance.

A core point is a DBSCAN observation with at least minPts points within ε distance, counting the observation itself. It meets the minimum density threshold and forms a building block of a cluster. section 2.5.2

Determine whether an observation is core by checking:

• The neighborhood radius ε. section 2.5.2
• The minimum point count minPts, including the observation itself. section 2.5.2
• Whether the neighborhood satisfies the density requirement; otherwise, classify the observation as a border point or outlier. section 2.5.2

DBSCAN parameter choices

A DBSCAN run can use eps = 0.0375 and min_samples = 6, or eps = 1.5 and min_samples = 5. These values determine which observations qualify as core points. section 2.5.2

Figure 2.29 Core points are shown as squares; border points are shown as filled circles, while noise is unfilled circles. Together, these three are the building blocks for DBSCAN clustering.

Use the figure to distinguish core points from border points and noise in a DBSCAN cluster. section 2.5.2

How Data Without Labels covers it

How deep it goes

Defines the ε and minPts test and links it to density, but does not detail the full clustering procedure.

Examples

Illustrates core points with Figure 2.29 and DBSCAN configurations using alternative eps and min_samples values.

What it assumes

Assumes the neighborhood concept, then defines ε and minPts before classifying observations.

What you can do after

Evaluate whether observations satisfy a DBSCAN neighborhood density threshold.

How it connects

Connects core points with DBSCAN clustering, epsilon, minPts, border points, outliers, and density-connected points.

liveBookmanning.com

Read Data Without Labels on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  Advanced Algorithms and Data StructuresA core point in DBSCAN is a data point whose neighborhood contains at least the required number of dataset points within the specified radius.

Compare how 2 books explain core point →

Related concepts

• DBSCANused by · DBSCAN is a density-based clustering algorithm that groups observations in dense regions and identifies observations in lower-density regions as noise or outliers.
• pointkind of · A point is a zero-dimensional location in a geometric or coordinate space represented by an ordered set of coordinate values.

This explainer was generated by Manning's knowledge graph from the text of Data Without Labels. The section links point to the parts of the book it draws on.