In this chapter, we covered a number of ways to deal with multivariate data to reduce the number of available dimensions in the means of artificially computed continuous variables and to identify underlying, latent, and similarly numeric variables. On the other hand, sometimes it's rather difficult to describe reality with numbers and we should rather think in categories.

The next chapter will introduce new methods to define data types (clusters) and will also demonstrate how to classify elements with the help of available training data.