Both k-means and k-medoids algorithms can perform effective clustering. (a)Explain the strength and weakness of k-means in comparison with the k-medoids algorithm. (b)Explain the strength and weakness of Partition clustering scheme in comparison with a hierarchical clustering scheme (such as AGNES).

Explanation:

A. K- medoids algorithms is more robust and vigorous where there is noise and  outliers than K-means, this is because outliers have less effect on mediod or other extreme values than mean, although processing it is more expensive than when compared to K-means method.

B. Partitioning based clustering is done by both K- means and K- mediods. The advantage of the partitioning is that previous cluster steps can be overrule through iterative relocation compared with hierarchical method which does not permit adjustments once a split or merge is carried out. This flaw in the hierarchical method can affect the quality of the resulting clustering.