Designing clothing that fit the population well poses an important challenge to industry, from mass market manufacturing to high-end designer labels. Clothes must fit the body elegantly and must be designed to hide obvious body flaws; otherwise the consumer will take his business elsewhere or the manufacturer will end up with old stock that does not sell.

This talk presents the results of a data mining exercise to obtain profiles of the typical consumer's bodies and demographics. To this end, we applied cluster analysis to a set of anthropometric body measurements in order to group the population into five clusters, each corresponding to a clothing size (small, medium, large, extra-large or extra-extra-large). We subsequently employed multi-view classification, in order to obtain sets of rules to learn which body measurements are of importance when designing clothes. Our results indicate that, for the different clothing sizes, the relevant anthropometric body measurements differ substantially. That is, different body measurements are used to characterize each clothing size. This information, together with the demographic profiles of the typical consumer, provides us with new insight into our evolving population.