A multivariate forecasting technique used to predict group membership from a linear combination of two or more interval dependent variables. Used in archaeology, as for example to develop classification functions for otoliths from fish which once lived in an archeological site in order to differentiate the species they belonged to.
