Created by W.Langdon from gp-bibliography.bib Revision:1.4420

@Book{frey:2002, author = "Clemens Frey", title = "Virtual Ecosystems - Evolutionary and Genetic Programming from the perspective of modern means of ecosystem-modelling", publisher = "Institute for Terrestrial Ecosystems, Bayreuth", year = "2002", volume = "93", series = "Bayreuth Forum Ecology", address = "Bayreuth, Germany", note = "(in German)", email = "frey@mathematik.tu-darmstadt.de", keywords = "genetic algorithms, genetic programming", ISSN = "0944-4122", URL = "http://www.bayceer.uni-bayreuth.de/bitoek/en/best/best/best.php?id_obj=9207", broken = "http://www.bitoek.uni-bayreuth.de/bitoek/en/pub/pub/pub_detail.php?id_obj=7556", size = "199 p.", abstract = "The realm of Evolutionary Computation covers many tools commonly used for solving complex discrete and continuous global optimization problems. These methods, which are known as Genetic Algorithms, Evolution Strategies, Evolutionary Programming and Genetic Programming, stem from efforts of modeling adaptive systems, from engineering and computer science. They are based on the idea of restating the Darwinian principles of natural evolution in algorithmic terms in order to get problem-solving methods for non-biological applications. Today Genetic Algorithms, Evolution Strategies and Evolutionary Programming mainly serve as mathematical techniques of numerical optimization. Genetic Programming likewise is an adaptation technique, but there is a different focus: based on evolutionary principles Genetic Programming enables us to automatically generate computer programs.The underlying hypotheses of this book is that the main point of natural, biological evolution is its data processing aspect. Evolution is seen as a certain way of processing individuals' and populations' genetic data. Referring to Evolutionary Computation there is a very interesting question now: Is it appropriate to employ Genetic Programming and similar algorithms in order to investigate natural evolution? Of course this means turning around the application profile of Evolutionary Computation, so where do we have to alter its algorithmic structure and the like? Finally, supposed there is a modified method, how do the results of both the classic algorithm and the modified variant compare to each other?In the first chapter we state the general notion of a search strategy. It may be a living being's policy of resource allocation, for example, but the notion covers optimization methods, too. A search strategy shall be defined in mathematical terms as being a dynamical system combined with a quality measure which is based on the trajectories the dynamical system produces. The author proposes a precise formulation for what a search strategy is generally claimed to accomplish, namely to generate dynamic behavior which gets us to the neighborhood of a predefined goal, possibly obeying certain constraints within the dynamics of the search process.Chapter two contains a gentle introduction into the field of Evolutionary Computation, namely Adaptive Systems, Genetic Algorithms, Evolution Strategies and Evolutionary Programming. We focus on Genetic Programming, however, and take a look at a paradigmatic experiment for automatically finding search strategies, i.e. the so-called artificial ant-experiment. In doing so the reader is also shown how the mathematical framework built in the first chapter may be used to formulate the artificial ant-problem.", abstract = "The following chapter addresses the issue of artificially creating evolution in virtual or simulated ecosystems and the question whether this can be done with the help of Evolutionary Computation. Since we want to analyse shortcomings of the conventional approaches and necessary adjustments, basic features of natural evolution are stated and discussed at first. Then we take a closer look at the area of Artificial Life and discuss specific software from this field. This discussion is concerned with so-called strong approaches like tierra and avida as well as weak approaches like the ecosystem-oriented Tragic++ system; besides, connections to social learning paradigms and Nouveau Artificial Intelligence are highlighted. Taking this broad view into account we conclude this chapter by listing a set of features which have to be comprised by a serious a model for evolution in virtual ecosystems. The gist of these desired features says that it is feasible to represent strategy programs as trees like in Genetic Programming, for this kind of representation causes a non-trivial, morphogenic mapping between the genotypic and the phenotypic space. It has to be conceded, however, that exogenously and a-priori given fitness-functions as well as the synchronous reproduction schemes which are almost always used in Genetic Programming are not appropriate for modeling evolution in virtual ecosystems. Chapters four to six describe how a system called MathEvEco was formulated and implemented according to these guidelines. Chapter four focuses on strongly typed tree representations of programs. Feasible sets of strongly typed program trees are defined precisely and their relationship with context-free grammars and the parameter-dependent evaluation of program trees are investigated in mathematical terms. These mathematical tools having been made available, genetic operators and initialization procedures of MathEvEco are stringently formulated in the fifth chapter. The system was supposed to be as flexible as possible. To this end the author has not only accessed a strongly typed version of the very classic crossover operator, but included a bunch of strongly typed mutation operators and the novel PTC2 algorithm for randomly generating program trees. In order to allow algorithmic comparisons the operators may be assembled in two fundamentally different ways; they may either be merged into a system of common Genetic Programming or they may be assembled as the desired system for modeling evolution in virtual ecosystems. Both of these possibilities are described, still in mathematical terms.The resulting systems are called MathEvEco-GP and MathEvEco-AL, respectively.", abstract = "While chapter five has been written in order to allow these systems to be communicated in a transparent and precise manner, chapter six shall illuminate their actual implementation within the scope of the mathematical software system Mathematica. To this end we show how program trees are handled in Mathematica, how model-specific and problem-specific knowledge is to be inserted by the user of MathEvEco, and in which way the various genetic operators have been implemented. Since MathEvEco can not only be run on a single machine but rather on clusters of workstations, there is a special treatment of aspects of parallel programming, too. Finally the functionality of MathEvEco is exemplified by means of a symbolic regression problem.The final chapter seven is dedicated to a case study. It consists of automatically generating search devices which is a special case of the general setting having been introduced in chapter one. There are a two different interpretations of this special problem. On the one hand side it may be understood in terms of numerical optimization; we presuppose an multi-modal objective function which may be imagined as a three-dimensional surface having many peaks. Strategies have to be evolved by MathEvEco-GP which are only provided with local information about this surface but are nevertheless required to lead the search devices to one of the highest peaks. On the other hand side the special problem may be understood in terms of an ecosystem where many organisms struggle for allocating BibTeX entry too long. Truncated

Genetic Programming entries for Clemens Frey