A. E. Eiben, R. Hinterding, and Z. Michalewicz, "Parameter Control in Evolutionary Algorithms", IEEE Transations on Evolutionary Computation, IEEE, 3(2), 1999, pp. 124-141.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....of automated control of evolution, in the context of efforts to minimize the time required to solve function optimization problems. Preliminary efforts were devoted to proving that automated control is feasible (e.g. 15, 2] and current research in this area is proceeding along several fronts [12]. Although the potential technological value of this work is obvious, it is unclear how, if at all, it will illuminate the theoretical questions about the evolution of evolvability. In particular, all the above cited work presumes that evolution is driven by a fixed and externally specified ....

Eiben, A. E., Hinterding, R., Michalewicz, Z. 1999. Parameter control in evolutionary algorithms. IEEE Transactions on Evolutionary Computation 3 (2), 124--141.

....is also needed to help ensure that the final population contains locally optimal points. Most real coded EAs perform local refinement with a mutation operator. In particular, adaptive methods that dynamically rescale step length parameters used for mutation have proven particularly e#ective [8]. Unfortunately, the analysis of adaptive real coded EAs has proven quite challenging. We argue that one of the reasons is that the formulation of common real coded EAs makes them di#cult to analyze. In particular, commonly used mutation operators generate new points from a continuous ....

A. E. Eiben, R. Hinterding, and Z. Michalewicz. Parameter control in evolutionary algorithms. IEEE Trans Evolutionary Computation, 3(2):124 -- 141, 1999.

....as well as any other search algorithm. The sensitivity of EAs to parameter values has lead to a wide variety of parameter control approaches. This makes it di#cult to suggest a good taxonomy. Eiben et al. suggest a taxonomy based on when (before or during the run) the parameters are determined [41]. However, parameters are seldomly determined without performing a few test runs, which makes the criterion when inadequate 49 for distinguishing between techniques. In my view, the key issue is whether or not the control method adapts to the search process. In non adaptive control, the ....

Eiben, A. E., Hintering, R., and Michalewicz, Z. (1999). Parameter Control in Evolutionary Algorithms. IEEE Transactions on Evolutionary Computation, 3(2):124--141.

....the problem parameters within the chromosome. Like co evolution, this is another means of providing a GA with more flexibility to conduct it s search effectively and allows the algorithm to adapt itself to the problem during a run [7, 8] For more information the reader can refer to Eiben et al. [5] where a survey of work in the field of parameter control is provided. Intuitively the two combined approaches of co evolution and parameter control should improve the performance of a GA. In this paper we propose a number of model variants of the CCGA 1 which utilise the GAVaPS adaptive ....

Eiben, A., Hinterding, R. and Michalewicz, Z. Parameter Control in Evolutionary Algorithms. In: IEEE Transactions on Evolutionary Computation Vol. 3 No. 2 124141 (1999)

....parameters value. These parameters are: i) N, number of candidate solutions; ii) g . maximum number of iterations; iii) Pc, crossover probability; iv) p, mutation probability. There is no analytical relationship between the performance and the population size N of a GA optimizer [7]. However, we do know the effects of the population size on the computation time of the algorithm: the larger is N the greater is the computation load. Thus, we need to find a balance between the coverage of the search space and the time it takes to complete g . iterations. In other to ....

A.E. Eiben, R. Hinterding, Z. Michalewicz, "Parameter control in evolutionary algorithms," 1EEE Transactions on Evolutionary Computation, vol. 3, no. 2, pp. 124-141, july 1999. - 731 -

....nature of the genetic algorithms and many researches have been performed on finding the optimum set of parameters for genetic algorithms. There is a very good survey of various forms of control, which have been studied in evolutionary computation community in recent years, with 144 references in [7]. Their classification covers the major forms of parameter control in evolutionary computation and suggests some directions for further research. We used a Message Passing Interface (MPI) based cluster computer [ASMA] for running our parallel annealed genetic algorithm. Message passing is a ....

....crossover and mutation rates) will not be proper for our problem. Some research is done on adaptively changing these parameters during the run of the program. Unfortunately, how to adapt the parameters to the current problem is another unsolved question and it is an important research area [7]. However, if we run our genetic algorithm with different crossover and mutation rates on each computer, we overcome the hardest problem and disadvantage of genetic algorithms. At least one of the parameter sets will be relatively suitable for the current problem. Choose the best GA 1 GA 2 GA 3 ....

Eiben, A. E., R. Hinterding, and Z. Michalewicz, "Parameter Control in Evolutionary Algorithms," IEEE Tr. on Evolutionary Computation, Vol. 3, No. 2, pp. 124-141, July 1999.

....as fast as possible. Some SAGAs adjust the mutation probability depending on time in order to prolong the evolution (i.e. to avoid the premature convergence) see e.g. 4] Some other use feedback from the population for that purpose [5] A good overview of different SAGA methods can be found in [3] and different desired behaviours of SAGA are described in [1] Spears and De Jong have explained with their dis ruption construction theory (recently summarized in [9] how mutation and crossover affect schemata. Although this theory gives interesting insight in behaviours of these operators, ....

A. E. Eiben, R. Hinterding, and Z. Michalewicz. Pa- rameter control in evolutionary algorithms. IEEE Transactions on Evolutionary Computation, 3(2), 1999.

....the following parameter settings: Population size: 30 Number of generation: 1000 Probability of crossover: 0.8 Probability of mutation: 0.007 The parameter settings were based on results of several preliminary runs. They are comparable to the typical values reported in the literature [1]. 3.3 Selection Mechanism The selection mechanism is responsible for selecting the parent chromosome from the population and forming the mating pool. The selection mechanism emulates the survivalof the fittest mechanism in nature. It is expected that a fitter chromosome receives a higher ....

A.E.Eiben, R.Hinterding, and Z.Michalewicz. Parameter control in evolutionary algorithms. IEEE Trans. on Evolutionary Computation, 3(2):124--141, 1999.

....is described by means of fixed kernels around the positions of individuals, the exploration density varies when individuals move on. But this does not quite capture what we actually meant by requiring variable exploration. Rather it is intuitive to call for adaptive codings . The review (Eiben, Hinterding, and Michalewicz 1999) (and also (Smith and Fogarty 1997) summarizes and classifies such approaches. Their discussion is based on the assumption that the coding (G, h, M G ) depends on some parameters x # X called strategy parameters; we write (G x , h x , M G x ) They classify di#erent approaches by ....

Eiben, A. E., R. Hinterding, and Z. Michalewicz (1999). Parameter control in evolutionary algorithms. IEEE Transactions on Evolutionary Computation 3, 124--141.

....in the optimal parameter set as the parameters are not independent and often interact in complex ways. The problem of finding optimal control parameters for genetic algorithms has been studied by many (De Jong 1975; 1980; Grefenstette 1986; Schaffer et al. 1989; Bramlette 1991; Wu Chow 1995; Eiben, Hinterding, Michalewicz 1999). One approach that has shown promise in several contexts is to use the genetic algorithm itself to search for good control parameter settings a so called meta level genetic algorithm approach. However, one drawback with this approach is its computational requirements. The process of executing ....

Eiben, A. E.; Hinterding, R.; and Michalewicz, Z. 1999. Parameter control in evolutionary algorithms.

....time such as the mutation rate or the variance of a Gaussian mutation operator. One idea to cope with this lack of generality is to use self adaptive algorithms where the EA parameters become a part of the simulated evolutionary process (e.g. Back, 1992] for a survey on parameter control see [Eiben et al. 1999]) Though this idea attracted attention in recent years, it dates back to the very beginning of the field, when Rechenberg introduced evolutionary strategies [Rechenberg, 1973] Here, the mutation rate becomes a part of the individual s genome. A recent promising idea in the context of ....

Eiben, A. E., Hintering, R., and Michalewicz, Z. (1999). Parameter control in evolutionary algorithms. IEEE Transactions on Evolutionary Computation, 3(2):124--141.

....across the subset size space. Hence, it enables more exploration to take place across a wider range of the non dominated front. Adaptive operator settings. General guidelines often suggest values of crossover no less than 0. 6 and mutation rate equal to 1 l c , where l c is the chromosome length [14]. When employing the crossover operator introduced in this paper, there is no need for adapting the crossover rate, while there is a simple way to control the mutation rate. 3 Subset Size Oriented Common Features Crossover Operator Common uniform or n point crossover operators can be disruptive ....

A. A. Eiben, Hinterding R. and Michalewicz, Z., (1999) "Parameter control in evolutionary algorithms," IEEE transactions on evolutionary computation, vol. 3, pp. 124-141.

....model is that of self adaptation. The issue of controlling values of various parameters of an evolutionary algorithm is one of the most important and promising areas of research in evolutionary computation: it has a potential of adjusting the algorithm to the problem while solving the problem [11]. The Patchwork model incorporates some self adaptation ideas [16, 25] as an agent carries additional chromosomes (apart from a solution chromosome) which determine its behavioral patterns; the values present in these additional chromosomes are self adaptive. The basic ideas behind the ....

Eiben, A.E, Hinterding, R., and Michalewicz, Z., Parameter Control in Evolutionary Algorithms, submitted for publication, 1999.

....current locations [32] ffi Adaptation and self adaptation mechanism. Dynamical adjustment of the algorithm to the non stationary environment is the next feature of the efficient optimization. So adaptive and self adaptive techniques are the next significant extension of evolutionary algorithm [1, 3, 8]. In adaptation the parameters of the algorithm are updated using statistic or heuristic rules to determine how to update. Update of the parameters in the genetic process in parallel with searching of the optimum is called a self adaptation. Both these techniques of parameters update were applied ....

Eiben, A., E., Hinterding, R., Michalewicz, Z., "Parameter Control in Evolutionary Algorithms ", Technical Report TR98-07, Department of Computer Science, Leiden University, Netherlands, 1998.

No context found.

A. E. Eiben, R. Hinterding, and Z. Michalewicz, "Parameter Control in Evolutionary Algorithms", IEEE Transations on Evolutionary Computation, IEEE, 3(2), 1999, pp. 124-141.

No context found.

A.E. Eiben, R. Hinterding, and Z. Michalewicz. Parameter control in evolutionary algorithms. IEEE Trans. on Evolutionary Computation, 3(2):124--141, 1999.

No context found.

A. E. Eiben, R. Hintering, and Z. Michalewicz. Parameter control in evolutionary algorithms. IEEE Trans. on Evolutionary Computation, 3(2):124--141, 1999.

No context found.

A. E. Eiben, R. Hinterding, and Z. Michalewicz. Parameter control in evolutionary algorithms. IEEE Trans. on Evol. Comp., 3(2):124--141, 1999.

No context found.

A. E. Eiben, R. Hinterding, and Z. Michalewicz. Parameter control in evolutionary algorithms. IEEE Transactions on Evolutionary Algorithms, 3(2):121--141, July 1996.

No context found.

A.E. Eiben, R. Hinterding, and Z. Michalewicz. Parameter control in evolutionary algorithms. IEEE Trans. on Evolutionary Computation, 3(2):124--141, 1999.

No context found.

A. E. Eiben, R. Hinterding, and Z. Michalewicz, "Parameter control in evolutionary algorithms," IEEE Trans. Evol. Comput., vol. 3, pp. 124--141, July 1999.

No context found.

A. E. Eiben, R. Hinterding, and Z. Michalewicz. Parameter control in evolutionary algorithms. IEEE Trans Evolutionary Computation, 3(2):124 -- 141, 1999.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC