R. M. Friedberg. A learning machine: I. IBM Journal of Research and Development, 2:2--13, 1958. 23  Home/Search   Document Not in Database   Summary   Related Articles   Check

....which, in turn, can be perceived as a search through a space of potential solutions. Since usually we are after the best solution, we can view this task as an optimization process. The first works on the use of evolution inspired approaches to problem solving date back to the late 1950s [13, 14, 28, 33, 34]. Independent and almost simultaneous research conducted by Rechenberg and Schwefel on evolution strategies [81, 82,86, 87] by Holland on genetic algorithms [44, 46] and by Fogel on evolutionary programming [31, 32] triggered the study and the application of evolutionary techniques. Three ....

R. M. Friedberg. A learning machine: I. IBM Journal of Research and Development, 2:2--13, 1958. 23

....which, in turn, can be perceived as a search through a space of potential solutions. Since usually we are after the best solution, we can view this task as an optimization process. The first works on the use of evolution inspired approaches to problem solving date back to the late 1950s [2, 3, 5, 8, 9]. Independent and almost simultaneous research conducted by Rechenberg and Schwefel on evolution strategies [31,32,34,35] by Holland on genetic algorithms [12,13] and by Fogel on evolutionary programming [6, 7] launched the study and the application of evolutionary techniques. Evolutionary ....

....XX, NO. Y, MONTH 2000 107 TABLE II Genome encoding of parameters for both species. Genome length for membership functions is 54 bits. Genome length for rules is 19 Theta Nr 1, where Nr denotes the number of rules. Species 1: Membership functions Parameter Values Bits Qty Total bits P [1 8] 3 9 27 d [1 8] 3 9 27 Total 54 Species 2: Rules Parameter Values Bits Qty Total bits A [0 3] 2 9 Theta Nr 18 Theta Nr C [1,2] 1 Nr 1 Nr 1 Total 19 Theta Nr 1 CoCo algorithm with the same evolutionary parameters for both species. Values and ranges of values used for these ....

[Article contains additional citation context not shown here]

R. M. Friedberg. A learning machine: I. IBM Journal of Research and Development, 2:2--13, 1958.

....Jefferson 90] is a good example. It demonstrates the evolution of both neural networks and finite state machines through a genetic algorithm running on a bit string representation. More conventional computer programs have also been processed with genetic algorithms, such as the pioneering work of [Friedberg 58] Friedberg et al. 59] and more recently that of [Ray 90]. Robot programs, and in particular behavior based robot programs, are much more complex than any programs that have been reported in the literature to have been so evolved. A reasonable comparison might be in terms of the memory taken to represent the programs. By this measure behavior based ....

"A Learning Machine, Part II", R. M. Friedberg, B. Dunham, and J. H. North, IBM Journal of Research and Development 3, 282-287.

....Jefferson 90] is a good example. It demonstrates the evolution of both neural networks and finite state machines through a genetic algorithm running on a bit string representation. More conventional computer programs have also been processed with genetic algorithms, such as the pioneering work of [Friedberg 58] Friedberg et al. 59] and more recently that of [Ray 90] Robot programs, and in particular behavior based robot programs, are much more complex than any programs that have been reported in the literature to have been so evolved. A reasonable comparison might be in terms of the memory taken to ....

"A Learning Machine, Part I", R. M. Friedberg, IBM Journal of Research and Development 2, 2-13.

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