Multi-set and multi-way data are collected in diverse scientific areas, like chemistry, metabolomics, signal processing and the social sciences. Multi-set data typically involve multivariate data that are organized in different sets or blocks, which have one mode in common. Such data arise when the observation units involved stem from different groups (i.e., blocks sharing the variable mode), or when multiple sorts of information are collected on the same observation units (i.e; blocks sharing the observation unit mode). Multi-way data require the data to be fully crossed, implying that the observation units are measured in different conditions on the same variables. Multi-set and multi-way models aim at capturing the intricate structure in those data sets, primarily using dimensional and categorical reduction models. These models also have applications in scientific computing, where the dimensional reduction is used to simplify computations.
The specialized team focuses on developments in multi-set and multi-way modeling, including their mathematical basis, algorithms and applications. It aims at stimulating the dialogue between the diverse communities interested in multi-set and multi-way modeling, such as statistics, tensor algebra, chemometrics, and psychometrics.