Summary

In this chapter, we learnt about clustering and we saw how this approach helps to group different items into groups with each group having items which are similar to them in some form. Clustering is an example of unsupervised learning and there are lots of popular clustering algorithms that are shipped by default in the Apache Spark package. We learnt about two clustering approaches, the first being k-means approach where items that are closer to each other based on some mathematical formula like Euclidean distance and so on were grouped together. We also learnt about bisecting k-means approach which is essentially and improvement on the regular k-means clustering and is creating by being a combination of hierarchical and k-means clustering. We also applied clustering on a sample dataset of retail from UCI. On this sample case study we segmented the customers of the website using clustering and tried to figure out the important customers for an online e-commerce store.

In the next...