Design and adaptation of algorithms
The technical problem determines the parameters of the real system that are to be computed, simulated or optimised. The solution generally requires several different numerical or combinatorial algorithms, such as the solution of nonlinear systems of equations or the processing of complex geometries.
In practise, these algorithms have to be adapted to the concrete case: A numerical method might work (converge) only in a certain region, or a combinatorial algorithm is efficient only for some classes of data. Often, problem specific information can be used to design efficient and robust ad-hoc algorithms.
Typically, the textbook version of a single method does not do, but a clever combination of algorithms from diverse branches of applied mathematics and computer science is required in order to achieve a satisfying solution.
CMM and its partners are experienced with a number of methods from diverse domains, their modification and the development of new algorithms, made to measure for a given task. We help you finding suitable approaches, investigate and improve already implemented algorithms, or to take over the algorithm development completely.