Hierarchical and k-Means Clustering

Example 6-6 introduced a greedy approach to clustering, but there are at least two other common clustering techniques that you should know about: hierarchical clustering and k-means clustering. Hierarchical clustering is superficially similar to the greedy heuristic we have been using, while k-means clustering is radically different. We’ll primarily focus on k-means throughout the rest of this chapter, but it’s worthwhile to briefly introduce the theory behind both of these approaches since you’re very likely to encounter them during literature review and research. An excellent implementation of both of these approaches is available via the cluster module that you can easy_install, so please go ahead and take a moment to do that.