Now that we have several methods for generating features from our data, we need a way to extract useful information from a vector of features. One possible approach is to take the set of feature vectors and examine if subsets of the feature vectors fall into clusters in an informative manner. For example, we could obtain feature vectors corresponding to different organisms and see if they cluster according to their phylogenetic relationships. Such clustering would suggest that the features are related to evolutionary relationships. Another situation would be where we have a set of data that we know corresponds to distinct conditions or classes, and we want to ascertain which class a new data point belongs to. This process is called classification, which we turn our attention to next.