Cluster analysis, or clustering, is a process of grouping a set of observations based on their similarities. The idea is that the observations in a cluster are more similar to one another than the observations from other clusters. Hence, the outcome of this algorithm is a set of clusters that can identify the patterns in the dataset and arrange the data into different clusters.
Clustering algorithms are referred to as unsupervised learning algorithms. Unsupervised learning does not depend on predicting ground truth and is designed to discover the natural patterns in the data. Since there is no ground truth provided, it is difficult to compare different unsupervised learning models. Unsupervised learning is generally used for exploratory analysis and dimensionality reduction. Clustering is an example of exploratory analysis. In this...
