T.Back and F.Hoffmeister and H.Schewefel
ESs are algorithms which imitate the principles of natural evolution as a method to solve parameter optimization problems. The development of ES from the first mutation-selection scheme to the refined (µ,lamda)-ES including the general concept of self-adaption of the strategy parameters for the mutation variances as well as their covariances are described.
Martin Pelikan, David E. Goldberg and Shigeyoshi Tsutsui
A method which combines competent genetic algorithms working in discrete domains with adaptive evolution strategies working in continuous domains is described. Discretization is used to transform solutions between the two domains.Experiments with Bayesian optimization algorithm are presented.
Garrison W. Greenwood and Christian Lang and Steve Hurley
Finding feasible schedules for tasks running in hard, real-time distributed computing systems is generally NP-hard. This paper describes a heuristic algorithm using evolutionary strategies. Our results indicate the evolutionary strategies can find feasible schedules (assuming they exist) in very short periods of time.
R. Kunz and R. Tetzla
In this paper a new learning algorithm for Cellular Neural Networks is presented based on evolutionary strategies. The proposed global optimization procedure isdiscussed in detail and the performance on various parameter determination problems will be shown too.
F. Schweitzer, W. Ebeling, H. Rosé and O. Weiss
The road optimization problem belongs to the class of frustrated optimization problems. In this paper, a special class of evolutionary strategies, such as the Boltzmann and Darwin and mixed strategies, are applied to find differently optimized solutions (graphs of varying density) for the road network in dependence on the degree of frustration.
A.M. Tyrrell, G. Hollingworth and S.L. Smith
This paper explores the possibilities of using evolutionary techniques to first produce a processing system that will perform a required function, and then consider its applicability for producing useful redundancy that can be made use of in the presence of faults, ie is it fault tolerant? Results obtained using Evolutionary Strategies to automatically create redundancy as part of the "design" process are given. The experiments are undertaken on a Virtex FPGA with intrinsic evolution taking place.
Werner Ebeling, Helge Rosé and Johannes Schuchhardt
The main elementary processes and strategies of evolution are investigated and described by simple mathematical models (stochastic networks). Special attention is devoted to Fisher-Eigen type models as well as to Boltzmann -, Darwin - and Haeckel - strategies modelling basic elements of frustrated problems in biological evolution respectively.
Helge Rose
To characterize the complexity of an optimization problem one may introduce a measure which remains invariant with regard to different schemes of representation: the density of states. It is the probability that an arbitrary chosen state has a certain fitness value. The knowledge of this probability makes it possible to estimate the optimal fitness value and the computational effort to find a better solution of the problem. A general method is presented which allows to approximate the density of states during the optimization process.