In k-means clustering, the results obtained are highly dependent on the initial allocation of observations to clusters.

Clustering is often an exploratory process that can aid in understanding the data.

Clustering is a supervised learning technique.

In k-means clustering, the number of clusters need to be pre-specified.