Within the general context of the study of multiway methods a number of specific subjects are currently under investigation:
Robustifying existing multiway methods (with Prof. M. Huber, Katholieke Universiteit, Leuven) Most methods for multiway analysis lean heavily on least-squares optimisation methods which are particularly sensitive for outliers. Studies are under way to evaluate this sensitivity and to propose to investigate more robust alternatives.
Applicability of multiway methods in brain and cognition studies. Typically fMRI studies involve multiway data, say voxels by time by conditions. Several models have been proposed to deal with them and the aim of this project is to investigate such proposals. This project is still in its infancy.
In the past years a major effort has taken place in developing user-friendly software for multiway analysis. A completely windows-driven programme is now in its final stages of testing and a first
version of a manual is completed. Honing both the programme and the documentation is the next step as well as marketing the software. In addition, preparations are under way to extend the present version with several other programmes which are already in existence but have not been included in the windows-driven interface.
Explicit efforts are made to publish signature analyses with multiway methods to make researchers aware of their potential.