For which type of data should K-means clustering be used?
K-means assigns observations to clusters by calculating distances to numeric centroids and updating those centroids with means. It therefore requires numeric data.
Community Discussion