Various clustering methods available
All clustering algorithms can be classified into four groups, as follows:
- Hierarchical clustering: Using this method, a number of clusters are prepared from the dataset, forming a hierarchy, and the clusters are then grouped using a tree structure so that the entire dataset can be represented as a single tree structure. The advantage of this method is that unlike partitioning-based clustering, we don't have to specify the number of clusters to be created from the dataset.
- Partitioning clustering: In this approach, the dataset is divided into a finite group of clusters based on a distance function in such a way that the within-cluster similarity is maximum and between-cluster similarity is minimum. K-means, which ...