This chapter covered the background and thought process that goes into designing a clustering algorithm for data mining work. It then introduced common clustering methods in the field and illustrated a comparison between all of them with toy datasets. After reading this chapter, you should know the difference between algorithms that cluster based on means separation, density, and connectivity. You should also be able to see a plot of incoming data and have some intuition on whether clustering fits your mining project. In addition, you should have a good idea of what method to try first.

The next chapter will cover common prediction and classification strategies, as well as introducing the concepts of loss functions, gradient descent, and cross validation.