Holdout set

This term is used in the context of multidimensional scaling procedures. In MDS, generally two sets of data are obtained. The holdout set is the second set of data that is used to assess reliability and validity.