DeJong, K. & Spears, W. 1993. On The State of Evolutionary Computation. In Proceedings of the Fifth International Conference on Genetic Algorithms Mining, 618-623. San Mateo, CA: Morgan Kaufmann.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....tasks to which AI has been applied that are inherently competitive, for example, games and auctions. In coevolution, the fitness of individuals in one species cannot be evaluated independently of the individuals in the other species. As the species evolve, the fitness landscape of each is changing[12]; this has been referred to as coupled fitness landscapes [13] A well known issue associated with the fitness evaluation of multiagent systems is credit assignment. When the system performs well or poorly which of the individual agents gets the credit or blame. For competitive coevolution ....

DeJong, K. & Spears, W. 1993. On The State of Evolutionary Computation. In Proceedings of the Fifth International Conference on Genetic Algorithms Mining, 618-623. San Mateo, CA: Morgan Kaufmann.

....(evaluate childs) ffl select the survivors from actual fitness (limit size of population) Genetic algorithms (GAs) owe their name to an early emphasis on representing and manipulationg individuals in terms of their genetic makeup rather than using a phenotypic representation. [JS93] The internal representation are binary strings, which are manipulated by recombination crossover and mutation operators. For GAs mutation is less important than recombination crossover, it is used to ensure diversity in the population during the optimization run. To be domain independent there ....

....They displace the exponentially diminishing schemata in bad (low fitness) bit strings. 5.2.3 Evolutionsstrategien Evolutionsstrategien . ESs) were developed with a strong focus on building systems capable of solving difficult real valued parameter optimization problems . [JS93] They walk (better diffuse) uphill on ways of steepest increase. Rec94] The internal representation are real valued vectors representing the parameters to be optimized. They are manipulated primarily by the mutation operator, which changes the entries according to a Gaussian distribution. ....

Kenneth De Jong and William Spears. On the state of evolutionary computation. In Genetic Algorithms, pages 618--623, University of Illinois, 1993. Morgan Kaufmann Publishers, Inc. ISBN 1-55860-2992.

.... seem to provide a sufficient account of innovation: how can new rules and conventions break the customary ones How should one reconcile stability and innovation Genetic algorithms bridge this hiatus by introducing mutation (for a discussion on different mechanisms for obtaining innovation, see DeJong and Spears, 1995) also the co learning algorithm introduced by Shoham and Tennenholtz, 1994) However, the view of innovation as accidental mutation does not do justice to the agents active role in the establishment of conventions. Agents representations and interpretations seem to have a fundamental part in ....

DeJong, K. and Spears, W. (1995) On the state of evolutionary computation, Proceedings of the Sixth Conference on Genetic Algorithms, 618-23.

....in the beginning of 1992 in a talk at the British Computing Society Workshop on Cellular Automata, held in London; this event did not have a full proceedings volume. Later on, it was accepted for Complex Systems 92, held in Australia, and published in the book that came out of it; that paper [de Oliveira 1993] is an abridged version of the original, and contains the essence of Chapter 4. The third version of the system, then named Enact, was first published in the technical report [de Oliveira 1992b] and later on accepted for a poster presentation and demonstration at the third Artificial Life ....

....is also worthy of mention. Goldberg 1989] is still the most accessible entry point to the field of genetic algorithms. The research pespectives in genetic algorithm as perceived in [De Jong 1985] are still very up to date, mainly if compared with an assessment of the field written nowadays, as in [De Jong and Spears 1993]. Of particular relevance to the artificial life community is the review presented in [Mitchell and Forrest 1993] and also the work on variable length genotypes presented in [Harvey 1994] Fogel 1992] traces back the history of evolutionary computation, specifically from the perspective of ....

[Article contains additional citation context not shown here]

Kenneth A. De Jong and W. Spears. On the state of evolutionary computation. In Stephanie Forrest, editor, Proceedings of the 5th International Conference on Genetic Algorithms, pages 618--623, San Mateo, CA, USA, 1993. Morgan Kaufmann.

.... processing and recognition, and engineering design [27, 87, 90] There are many issues in genetic algorithms that have been and continue to be studied, including issues of representations, population size, evaluation of fitness, selection of individuals for reproduction, and reproduction methods [88, 90,94, 95]. 3.3.2 Genetics based machine learning Genetics based machine learning is an extension of genetic algorithms (GAs) to machinelearning problems [27] This term is used in this thesis to cover the applications of the idea behind genetic algorithms to develop a heuristic or strategy that ....

K. DeJong and W. Spears, "On the state of evolutionary computation," in Proc. Fifth Int. Conf. Genetic Algorithms, Int. Soc. for Genetic Algorithms, June 1993, pp. 618-- 623.

....into the competition among structures, and this is readily done in the nGA format. The nGA model also holds promise for explorations in artificial life, particularly as concerns group interaction. Non traditional operators may prove attractive for use at upper levels of an nGA based Alife model. DeJong and Spears (1993) pointed out that, as evolutionary computation moves forward, there still are many potentionally important dimensions for EC research. We believe that the nGA model and the first realization under it, the DAGA2 implementation, have powerful and unique features as adaptive GA s. They appear to be ....

DeJong, Kenneth and William Spears (1993). "On The State of Evolutionary Computation," Proceedings of the Fifth International Conference on Genetic Algorithms, Morgan Kaufman, New York.

....search can be regarded as guided by a decision theory heuristic. Similarly, implicit parallelism results [2, 3] give indication of the effectiveness of the computation. The convergence of genetic algorithms is one of the most challenging theoretical issues in the evolutionary computation area [8, 18]. Several researchers have previously explored this problem mainly from two different perspectives: Analyses using Markov chains [5, 9, 14, 26] and analyses to identify characteristics of the objective functions that affect the behavior of genetic algorithms [13, 16] Goldberg and Segrest [14] ....

K.A. DeJong and W.M. Spears. On the state of evolutionary computation. In S. Forrest, editor, Proceedings of the Fifth International Conference on Genetic Algorithms, pages 618--623, 2929 Capus Drive, San Mateo CA 94403, 1993. University of Illinois at UrbanaChampaign, Morgan Kaufmann Publishers.

....are often applied to problems where there is a best solution (to which one might hope the GA under study will converge) there are many domains where convergence to the optimal string is neither desirable nor sensible. Evolutionary computation is increasingly moving towards dynamic environments [7]. In such environments the optimal string is not a static entity. If we wish to address difficult problems in complex dynamical settings, our algorithms must evolve to allow populations to remain heterogeneous even when some individuals are extremely successful at a particular time and ....

DeJong, K. & Spears, W. (1993). On the State of Evolutionary Computation. Proceedings of the Fifth ICGA, 618-623. Kaufmann, San Mateo, CA.

No context found.

K. De Jong and W. Spears, "On the State of Evolutionary Computation", Proc. ICGA '93, pp 618623

No context found.

K.A. DeJong and W.M. Spears. On the state of evolutionary computation. In S. Forrest, editor, Proceedings of the Fifth International Conference on Genetic Algorithms, pages 618--623, 2929 Capus Drive, San Mateo CA 94403, 1993. University of Illinois at UrbanaChampaign, Morgan Kaufmann Publishers.

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