Multi Way Classification

In this Chapter, the state-of-the-art approaches for the classification of multi-way data is presented and discussed. The theoretical basis and applicative guidelines for multilinear (or multi-way) Partial Least Squares Discriminant Analysis (NPLS-DA) and Multi-way Soft Independent Modelling of Class Analogy (NSIMCA) are detailed. Furthermore, two-dimensional linear discriminant analysis (2DLDA) and a proposal for truly multilinear discriminant analysis are illustrated. The truly multi-way methods are compared to unfolding and feature extraction followed by bilinear classification. Practical hints are depicted through discussion of a case of study.