Three-way analyses:Problems and prospects

Three-way methods are multivariate data analysis tools that compress and visualize simultaneous variation of combinations of variables and objects. Unfolding, Tucker and PARAFAC models are examples of such methods. The interest in such tools is increasing in chemometrics. The history of three-way methods is sketched and the basic theory is presented, together with chemical applications. Problems in the three-way analyses area are summarized. Some prospects for future use of three-way methods are given.