K-Means Clustering in R Programming

K Means Clustering is an unsupervised learning algorithm that groups data into clusters based on similarity. This algorithm divides data into a specified number of clusters, assigning each data point to one. It is used in various fields like banking, healthcare, retail and media. In this article we will implement K Means Clustering in R Programming Language.

Working of K-Means Clustering Algorithm

K Means Clustering is an iterative algorithm that divides data into K clusters, aiming to group data points that are similar to each other while minimizing the distance between them and their cluster's centroid. Below is a step-by-step breakdown of how the algorithm works and how it assigns data points to clusters:

1. Choose the Number of Clusters (K): The number of clusters (K) is determined based on the dataset or domain knowledge. This value dictates how many clusters the algorithm will form.
2. Select Initial Centroids: K initial centroids are selected randomly from the dataset. These centroids represent the center of each cluster.
3. Assign Data Points to the Closest Centroid: Each data point is assigned to the nearest centroid by calculating the distance to each centroid. This creates K initial clusters where data points are grouped around the centroids.
4. Calculate New Centroids: After the initial assignment, the centroid of each cluster is recalculated. The new centroid is the average of all the data points assigned to that cluster.
5. Reassign Data Points: Data points are reassigned to the closest centroid after the centroids are updated. Some data points may switch clusters based on the new centroid positions. The process of recalculating centroids and reassigning data points is repeated until the centroids no longer change significantly.

K-Means Clustering in R

We are going to perform to K-Means clustering on the Iris dataset in R.

1. Importing the Dataset

We are loading the Iris dataset that contains measurements of Iris flowers' sepal and petal.

Output:

2. Installing Necessary Packages

We are installing and loading the required packages for clustering and visualization. We are using "factoextra" for cluster visualization and "cluster" for clustering operations.

3. Visualizing the Data

We are visualizing the Iris dataset using Sepal Length and Sepal Width to understand the distribution of data points.

Output:

4. Performing K-Means Clustering

We are applying the K-Means clustering algorithm to the dataset and setting the number of clusters to 3 (corresponding to the 3 Iris species).

Output:

5. Creating a Confusion Matrix

We are creating a confusion matrix to compare the actual species labels with the clusters predicted by K-Means.

Output:

6. Visualizing the Clusters

We are visualizing the clusters formed by K-Means using the factoextra package. This plot shows the clusters along with Sepal Length and Sepal Width.

Output:

This is a plot showing the three K-Means clusters based on Sepal Length and Sepal Width.