Regarding outliers, the HIPAMIMO algorithm finds 82 while HIPAMMO finds 92. Forty of them are considered outliers by both algorithms. Table 5 shows the percentage of outliers obtained by both algorithms for each class. Fig. 4 shows the scatter plots of bust vs. hip, together with the outliers according to HIPAMMO and HIPAM- IMO. We observe that HIPAMIMO only identifies outliers in the four bust classes corresponding to small and large sizes. Clothing industry practice for the mass production of clothing is to optimize sizes by addressing only the most profitable. Extreme sizes are usually offered as “special sizes” by companies focusing on this target market. Thus, results provided by HIPAMIMO are quite well aligned to this aim.Some other scatter plots showing medoids and outliers returned by HIPAMMO and HIPAMIMO can be found in the Supplementary Material. In addition, a comparison between the outliers detected by our algorithms and the outliers detected by a common method used in the apparel sizing literature to the same end and an R package that focuses on identi- fying multivariate outliers are also given in the Supplementary Material.