After using k-means to split the iris samples into three clusters, it is important to evaluate the quality of this grouping. A first approach is to compare the clusters with the iris species. This method then makes it possible to tackle the evaluation of a grouping's quality without requiring the samples to have been classified by species beforehand. This evaluation can also help choose the number of clusters best suited to the data. Iris: clusters and species To check whether the three clusters of iris samples match the species, it is possible to build a table showing their correspondence. This table has one column for
The rest of this lesson is reserved for subscribers Create your free account to try the sample exercises, then upgrade to full access whenever you like.