Analysis, Specification and
Development of Hybrid Algorithms for Optimization and
Satisfability Problems (AEDRHOS::unican).
The goal of this project is to develop new formal
techniques and computational tools for optimization and
satisfiability problems paying special emphasis on those of combinatorial
flavour. We shall focus our attention on the study and implementation of new
strategies to successfully tackle formal problems which may be extended to
practical situations from industry. We propose to combine the use of several
metaheuristics (such as evolutionary algorithms, state space search,
adaptive greedy algorithms or local search) with other techniques coming
from the fields of Mathematics and Computational Intelligence with the
purpose of generating new hybrid methodologies to deal with complex problems
and to extend and improve other existing ones. On the one hand, our aim is
to transfer results from the corpus of Mathematics (such as linear algebra,
numerical calculus, probability and approximation) and Computational
Learning Theory to the field of heuristics and evolutionary computation. By
the other hand, we seek to adapt and adjust these general methods to real
problem solving, to enable technology transfer to society. This double
transfer is in the nucleus of project AEDRHOS.
Algoritmos No Universales y Algoritmos Alternativos en
Eliminación Geométrica: un Estudio de Eficacia.