DBSCAN clustering in ML Python - Cheat Sheet & Quick Revision

Recall & Review

beginner

What does DBSCAN stand for in clustering?

DBSCAN stands for Density-Based Spatial Clustering of Applications with Noise. It groups points that are closely packed together and marks points in low-density regions as noise.

beginner

What are the two main parameters of DBSCAN?

The two main parameters are eps (epsilon), which defines the radius to search for neighbors, and min_samples, the minimum number of points required to form a dense region (cluster).

beginner

How does DBSCAN treat noise points?

DBSCAN labels points that do not belong to any cluster because they are in low-density areas as noise or outliers. These points are not assigned to any cluster.

intermediate

What is a core point in DBSCAN?

A core point is a point that has at least min_samples points (including itself) within its eps radius. Core points are inside dense regions and help form clusters.

intermediate

Why is DBSCAN good for clusters of arbitrary shape?

DBSCAN can find clusters of any shape because it groups points based on density, not on distance to a center. This allows it to detect clusters that are not round or evenly sized.

What does the parameter eps control in DBSCAN?

A The radius to search for neighboring points

B The minimum number of clusters

C The maximum number of iterations

D The distance between cluster centers

In DBSCAN, what happens to points that are not in any dense region?

A They form a separate cluster

B They are labeled as noise

C They are merged into the nearest cluster

D They are ignored during clustering

Which of these is NOT a characteristic of DBSCAN?

A Can find clusters of arbitrary shape

B Uses density to form clusters

C Identifies noise points

D Requires specifying the number of clusters beforehand

What is a border point in DBSCAN?

A A point with fewer than <em>min_samples</em> neighbors but within <em>eps</em> of a core point

B A point outside all clusters

C A point with more than <em>min_samples</em> neighbors

D A point that starts a new cluster

Which scenario is DBSCAN especially useful for?

A When clusters are spherical and well separated

B When the dataset is very small

C When clusters have varying shapes and sizes

D When you want to predict labels for new data

Explain how DBSCAN forms clusters and handles noise in simple terms.

Describe the roles of the parameters eps and min_samples in DBSCAN clustering.