Recall that the goal is to group together “similar” data – but what does this mean? No single answer – it depends on what we want to find or emphasize in the data; this is one reason why …

Once a Clustering has been obtained, it is important to assess its validity! The questions to answer: Did we choose the right number of clusters? Are the clusters compact? Are the …

Gain insight into the structure of the data. Find prototypes in the data. Until now we have assumed that the training data is labeled. Now we look at what can we do with data when we have not …

the goal of clustering is to determine the intrinsic grouping in a set of unlabeled data. But how to decide what constitutes a good clustering? It can be shown that there is no absolute “best” …

-Divisive clustering starts with all the objects grouped in a single cluster. Clusters are divided or split until each object is in a separate cluster. Agglomerative methods are commonly used in …

In general, this is a unsolved problem. However there are many approximate methods. In the next few slides we will see an example. For our example, we will use the familiar …

However, in practice, it’s a bit less clear: there are many ways of influencing the outcome of clustering: number of clusters, similarity measure, representation of documents.