You are on page 1of 3

2/14/23, 12:24 PM What Is the Difference Between Hierarchical and Partitional clustering?

| by Jason LZP | Medium

You have 2 free member-only stories left this month.


Sign up for Medium and get an extra one

Jason LZP Follow

Nov 16, 2019 · 3 min read · · Listen

Save

What Is the Difference Between Hierarchical


and Partitional clustering?

https://lzpdatascience.medium.com/what-is-the-difference-between-hierarchical-and-partitional-clustering-edc0d488c7c4#:~:text=An example of … 1/5


2/14/23, 12:24 PM What Is the Difference Between Hierarchical and Partitional clustering? | by Jason LZP | Medium

Clustering is an unsupervised learning technique which is exploratory in nature


and does not have a defined target or output. Clustering is often done to undercover
hidden patterns within a dataset or for real-world uses such as market
segmentation.

In this post, I would be mainly discuss about the difference between Hierarchical
and Partitional clustering.

What Is Hierarchical Clustering?


Hierarchical agglomerative clustering is one of the main types of clustering
algorithm and works using the following steps:

1. Each item within a dataset starts as an individual cluster (AKA: singletons)

2. The algorithm computes the proximity amongst each cluster

3. It then proceeds to merge each pair of closest clusters and computes the new
proximity amongst remaining clusters

4. This process is then repeated until there is one cluster left

A typical representation of Hierarchical Clustering is represented using a


Dendrogram

https://lzpdatascience.medium.com/what-is-the-difference-between-hierarchical-and-partitional-clustering-edc0d488c7c4#:~:text=An example of … 2/5


2/14/23, 12:24 PM What Is the Difference Between Hierarchical and Partitional clustering? | by Jason LZP | Medium

Example of a Dendrogram

What Is Partitional Clustering?


On the other hand, partitional clustering is another primary type of clustering
technique. The K-Means algorithm is one of the most popular partitional clustering
methods.
Open in app Sign up Sign In

The K-Means clustering requires the analyst to define K number of clusters or the
number of iterations before running the algorithm. As such, it relies heavily on the
analyst’s knowledge to classify the clusters in a meaningful way.

The K-Means clustering works using the following steps:

1. The analyst selects the number of K clusters to be used as initial centroids or the
number of iterations for the algorithm to run (AKA: Stopping criterion)

2. Objects closest to the centroid are grouped to form K number of clusters

3. With every iteration, the centroid distance for each cluster shifts and is updated

https://lzpdatascience.medium.com/what-is-the-difference-between-hierarchical-and-partitional-clustering-edc0d488c7c4#:~:text=An example of … 3/5

You might also like