In this chapter, we presented the K-means clustering algorithm and clustering ensemble methods. We explained how majority voting can be used in order to combine cluster assignments from an ensemble, and how it can outperform the individual base learners. Furthermore, we presented the OpenEnsembles Python library, which is dedicated to clustering ensembles. The chapter can be summarized as follows.
K-means creates K clusters, and assigns instances to each cluster by iteratively considering the cluster center to be the mean of its members. It can be sensitive to the initial conditions, and the selected number of clusters. Majority voting can help to overcome the algorithm's disadvantages. Majority voting clusters together instances that have a high co-occurrence. Co-occurrence matrices show how frequently a pair of instances has been assigned to the same cluster by...
