Learning Data Mining with Python - Second Edition

Evaluating the results

Clustering is mainly an exploratory analysis, and therefore it is difficult to evaluate a clustering algorithm's results effectively. A straightforward way is to evaluate the algorithm based on the criteria the algorithm tries to learn from.

If you have a test set, you can evaluate clustering against it. For more details, visit http://nlp.standford.edu/IR-book/html/htmledition/evaluation-of-clustering-1.html

In the case of the k-means algorithm, the criterion that it uses when developing the centroids is to minimize the distance from each sample to its nearest centroid. This is called the inertia of the algorithm and can be retrieved from any KMeans instance that has had fit called on it:

 pipeline.named_steps['clusterer'].inertia_ ...