Data Structure
Networking
RDBMS
Operating System
Java
MS Excel
iOS
HTML
CSS
Android
Python
C Programming
C++
C#
MongoDB
MySQL
Javascript
PHP
Physics
Chemistry
Biology
Mathematics
English
Economics
Psychology
Social Studies
Fashion Studies
Legal Studies
- Selected Reading
- UPSC IAS Exams Notes
- Developer's Best Practices
- Questions and Answers
- Effective Resume Writing
- HR Interview Questions
- Computer Glossary
- Who is Who
What are the types of Clustering in data mining?
There are various types of clustering which are as follows −
Hierarchical vs Partitional − The perception between several types of clusterings is whether the set of clusters is nested or unnested, or in popular terminology, hierarchical or partitional. A partitional clustering is a distribution of the group of data objects into non-overlapping subsets (clusters) including every data object is in truly one subset.
It can allow clusters to have subclusters, therefore it is required hierarchical clustering, which is a group of nested clusters that are assigned as a tree. Every node (cluster) in the tree (except for the leaf nodes) is the union of its children (subclusters), and the root of the tree is the cluster including all the objects.
Exclusive vs overlapping vs Fizzy − The clustering is all exclusive, as they create each object to an individual cluster. There are several positions in which a point can be located in higher than one cluster, and these situations are superior addressed by non-exclusive clustering.
In this method, an overlapping or non-exclusive clustering can follow the fact that an object can belong to higher than one group (class). For example, a person at a university can be both an enrolled candidate and an employee of the university.
In a fizzy clustering, each object applies to each cluster with a membership weight that is between 0 (categorically doesn't applies) and 1 (categorically applies). In other terms, clusters are considered as fizzy sets.
Complete versus Partial − A complete clustering creates each object to a cluster, whereas a partial clustering does not. The reason for partial clustering is that some objects in a data set cannot belong to clear groups. Several times objects in the data set can define noise, outliers, or "uninteresting background." For instance, some newspaper stories can share a common design, including global warming, while different stories are more universal or one-of-a-kind.
Therefore, it can discover the important topics in last month's stories, it is required to search only for clusters of documents that are hardly connected by a common theme. In some cases, a whole clustering of the objects is acquired. For instance, an application that needs clustering to organize files for browsing is required to guarantee that all files can be browsed.
973 Views
