Emerging computational techniques (ECT)

Organisation profile:The research group Emerging Computational Techniques (ECT) started up in 2001 with as main research objectives the development of new computational methods, in particular methods based upon interval arithmetic and symbolical-numeric techniques. Interval arithmetic is an essential tool in the process of validating scientific computations, since it automatically computes an upper- and lower bound for the correct result, rather than giving a more or less accurate approximation of the exact result. In an age in which computer systems are increasingly relied on, the validation of those systems should increase accordingly. In addition to their importance as a validation tool, interval techniques are an essential tool in the solution of certain optimisation problems, specifically in problems where one is interested in global maxima (minima) as well as in constraint satisfaction problems in combinatorical optimisation. Symbolic-numeric techniques are playing an increasingly important role in applications where computations with polynomials are essential, such as robotics, vision, computational biochemistry, ... In those applications the need is felt to revise existing symbolic algorithms, to be able to take errors in coefficients into consideration. Those errors change the nature of the questions traditionally being asked about polynomials, and as a consequence, also the algorithms needed to answer those questions. As is clear from the publications and the research projects of the group, in the past 6 years the emphasis has mainly been on the first of these two research tracks. More precisely, several techniques have been developed to validate results of scientific computations or to solve specific numerical problems in a validated way. The activities of the research group are of theoretic nature as well as implemented in software and applied to practical problems. Among these applications one finds the reconstruction of 2- and 3-dimensional shapes from moment information, the modelling of multi-parameter electromagnetic phenomena, and the reliable evaluation of several special functions.

Keywords:SYMBOLIC-NUMERIC COMPUTATION, FLOATING POINT, FUNCTION APPROXIMATION, SIMULATIONS, MODELLING, INTERVAL ARITHMETIC, INVERSE PROBLEMS

Disciplines:Applied mathematics in specific fields, Computer architecture and networks, Distributed computing, Information sciences, Information systems, Programming languages, Scientific computing, Theoretical computer science, Visual computing, Other information and computing sciences