Table 5.

Hierarchical clustering vs. k-Means clustering in miRNA data analysis with relevance in ovarian cancer.

Hierarchical Clustering	k-Means Clustering
Can’t handle large miRNA expression data—quadratic complexity	Can handle large miRNA expression datasets—linear complexity
Reproducible as every miRNA expressed is assigned a cluster, and the clustering occurs based on the closeness of previously generated clusters.	Unreproducible clustering due to the prerequisite of a random number of clusters.
Produces more intuitive results in the form of a dendrogram.	Produces less intuitive results if data does not group into hyper spherical clusters.
Poor performance and higher time of execution as the number of generated clusters increases.	Higher time of execution associated with large miRNA expression datasets.