Clustering is used to isolate and focus the analysis on the data points that lie
outside the norm. By identifying outlier data points, we subject them to additional
scrutiny with diagnostics not suitable for examining large numbers of points. For
details see the paper by Negres et al., pp. D12-D20.