Multiway and multiset data analysis

Abstract

The field of multiway data analysis is maturing and deserves a second special issue after the first one (Journal of Chemometrics, Vol. 23, Nur. 7 and 8). Also it is increasingly being recognized that multiway methods can be considered a special case of a more general class of methods, namely multiset or multiblock methods. We therefore present a set of 16 papers devoted to the topics of multiway and multiset data analysis. A major part of these papers were presented and discussed in the TRICAP (ThRee-way methods In Chemistry And Psychology) meeting held in Vall de Nuria - Spain, June 14th to June 18th, 2009 (http://www.tricap2009.info/, see also Chemometrics and Intelligent Laboratory Systems 99 (2009) 161). This is a special thematic issue on the topic of multiway and multiset data analysis and not an ordinary proceedings issue. Authors have been invited to contribute according to their previously known work in the field and papers were then submitted to the ordinary peer review system for their improvement. We have prepared this special issue with the goal of contributing significantly to the extension of these important topics in chemometrics and also in other related fields. Multiset data is a collection of data sets related to one central question. In most cases, it is not sufficient to analyze the data sets separately but a simultaneous analysis is required. Multiset data have a less rigorous structure than multiway data and it is not trivial how to analyze such data. One of the key notions is shared modes between the data sets, e.g., when multiple sets of variables are measured for the same samples. Simultaneous analysis then makes use of this shared sampling mode. The extension of multiway to multiset data analysis has also obvious and important implications for the models designed to investigate them, requiring alternative modelling approaches complementary to the powerful multilinear models which are already rather mature. Both theoretical and new applications of multiway and multiset data analysis methods are presented.