Related Articles

Related Articles

Difference between CURE Clustering and DBSCAN Clustering
  • Last Updated : 31 Dec, 2019

Clustering is a technique used in Unsupervised learning in which data samples are grouped into clusters on the basis of similarity in the inherent properties of the data sample. Clustering can also be defined as a technique of clubbing data items that are similar in some way. The data items belonging to the same clusters are similar to each other in some way while the data items belonging to different clusters are dissimilar.

CURE (Clustering Using Representatives) and DBSCAN (Density Based Spatial Clustering of Applications with Noise) are clustering algorithms used in unsupervised learning. CURE is a hierarchial based clustering technique and DBSCAN is a density-based clustering technique.

These are some differences between CURE and DBSCAN :

S.No. CURE Clustering DBSCAN Clustering
1. CURE Clustering stands for Clustering Using Representatives Clustering. DBSCAN Clustering stands for Density Based Spatial Clustering of Applications with Noise Clustering.
2. It is a hierarchial based clustering technique. It is a density based clustering technique.
3. Noise handling in CURE clustering is not efficient. Noise handling in DBSCAN clustering is efficient.
4. Algorithm:

  • Draw a random sample.
  • Partition the random sample.
  • Partially cluster the partition.
  • Outliers are identified and eliminated.
  • The partial clusters obtained are clubbed into clustered.
  • Label the result on storage.
Algorithm:



  • All the data sample points are labelled as core points, border points or noise points.
  • The noise points are eliminated.
  • All the core points are connected which lie under the vicinity of Eps of each other.
  • The core points which are connected to each other are grouped into a separate cluster.
  • Border points are assigned to each clusters.
5. It can take care of high dimensional datasets. It does not work properly for high dimensional datasets.
6. Varying densities of the data points doesn’t matter in CURE clustering algorithm. It does not work properly when the data points have varying densities

CURE Architecture:

DBSCAN Architecture:

Eps : Radius of circle
minPts : It is the minimum no. of points that must exist in the vicinity of eps.

machine-learning

My Personal Notes arrow_drop_up
Recommended Articles
Page :