Question

You have performed an unsupervised k-means clustering on a data set with two attributes and the results indicate a k of 2. Later, you determine the class values for each data instance (there are four class values) and a supervised clustering results in a k of 4. Provide a possible explanation for why the two clustering methods disagree on a k value and a draw a sketch of the two clusterings to go along with your explanation.

Answer 1

The below-shown scenario is an example of what can make such a clustering case,

Supervised Clustering is the problem of training a clustering algorithm to produce desirable clusterings: given sets of items and complete clusterings over these sets, we learn how to cluster future sets of items. Clustering algorithms accept a set of items and produce a partitioning of that set.

Whereas in the case of Unsupervised algorithm, we don't have any guidance we have a change possible, the reason is that we have the goal of this unsupervised machine learning technique is to find similarities in the data point and group similar data points together.

Now when we have a set of items let say of four categories in the supervised clustering then we compare the similarities, of each unclassified element to all the four clusters and thus we have 4 clusters.

Whereas as you can see in the picture we can have the same set of elements without any guidance, the key thing is here there is no particular way of separating the two clusters like blue and red similarly light blue and black.

So when we, have the case of the different algorithms we can have different clusters, actually, even with two unsupervised algorithms we can have the separate number of cluster and that is because it all depends on the similarity and difference measure.

If there is any doubt, you can ask in comments.