A road map for multi-way calibration models

• Published in: Analyst (London) 2017-08, Vol.142 (16), p.2862-2873
• Main Authors: Escandar, Graciela M, Olivieri, Alejandro C
• Format: Article
• Language: English
• ISSN: 0003-2654; 1364-5528
• DOI: 10.1039/c7an00822h

A large number of experimental applications of multi-way calibration are known, and a variety of chemometric models are available for the processing of multi-way data. While the main focus has been directed towards three-way data, due to the availability of various instrumental matrix measurements, a growing number of reports are being produced on order signals of increasing complexity. The purpose of this review is to present a general scheme for selecting the appropriate data processing model, according to the properties exhibited by the multi-way data. In spite of the complexity of the multi-way instrumental measurements, simple criteria can be proposed for model selection, based on the presence and number of the so-called multi-linearity breaking modes (instrumental modes that break the low-rank multi-linearity of the multi-way arrays), and also on the existence of mutually dependent instrumental modes. Recent literature reports on multi-way calibration are reviewed, with emphasis on the models that were selected for data processing. A road map is proposed for the selection of a multi-way calibration model according to the data properties.