Bedau, M. A. 1995. Three illustrations of artificial life's working hypothesis. In: W. Banzhaf and F. H. Eeckman (Eds.), Evolution and Biocomputation: Computational Models of Evolution. Springer, Berlin, pp. 53--68.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....of the algorithm, but the approach is highly problem specific. There has also been some interest in applying mathematical models used in the study of biological genetics to evolutionary optimisation, for example, Muhlenbein in (Muhlenbein 1994; Muhlenbein and Schlierkamp Voosen 1995) and Bedau in (Bedau 1995). Recently, Cross has given an accurate theoretical account of the search properties of the genetic algorithm for inexact matching (Cross 1998) While general theoretical models are beneficial in understanding the processes at play in genetic algorithms, their use in practical settings is limited. ....

M. Bedau (1995). Three illustrations of artificial life's working hypothesis. Lecture Notes in Computer Science 899, 53--68.

....method is Bedau and Packard s evolutionary activity statistics. We also measure system diversity, D, which is simply the number of different genotypes present in a system at a given time. Detailed definitions and motivations for evolutionary activity statistics are readily available elsewhere [2, 1, 3, 4, 13]. Evolutionary activity statistics aim to identify evolutionary innovations (here, new genotypes) that persist and continue to play a significant role in a system because of their adaptive value. These statistics fall into two broad classes: those reflecting evolutionary activity s extent and ....

Bedau, Mark A. 1995. Three illustrations of artificial life's working hypothesis. In W. Banzhaf and F. Eeckman, (Eds.), Evolution and biocomputation: computational models of evolution, (pp. 53--68). Berlin: Springer.

....where molecules collide and interact to produce new molecules [3] It can be called an artificial chemistry, because the interaction rules are defined by abstract algorithms. Through this abstraction one tries to extract the logic of information processing rather than simulating physical details [5]. The system consists of the following three components: 1) A soup (population) of objects. These objects may be character sequences [11] lambda expressions [8] binary strings [18, 2] or numbers. Here, we use binary strings with a constant length of 32 bit. 2) A collision or reaction rule. ....

M. A. Bedau. Three illustrations of artificial life's working hypothesis. In Wolfgang Banzhaf and Frank H. Eeckman, editors, Evolution and Biocomputing, Volume 899 of Lecture Notes in Computer Science, pages 53--68. Springer, Berlin, 1995.

....in large part because of the lack of objective and feasible ways to measure the relevant aspects of the adaptive dynamics of evolving systems. In the attempt to address this sort of problem, Bedau and Packard have defined statistics that measure the evolutionary activity of evolving systems [2, 1, 3], where these statistics are based on the assumption that, on average, the persistence of a component (e.g. genotype) in an evolving system reflects the component s continual adaptive significance, i.e. its continual adaptive advantage over its competitors. In the present paper we use these ....

....on many levels in most evolving systems. We define an innovation as the introduction of a new component into the system. In the case of this study of ECHO, the components we choose to study are entire genotypes. Previous studies have measured innovations on the level of individual alleles [2, 1], genotypes [3] and taxonomic families in the fossil record [3] 2. How should a given innovation contribute to the evolutionary activity of the system We measure activity contributions by attaching a counter to each component of the system. In the work we present here a component s activity ....

[Article contains additional citation context not shown here]

Bedau, Mark A., "Three Illustrations of Artificial Life's Working Hypothesis," Evolution and Biocomputation--- Computational Models of Evolution, W. Banzhaf and F. Eeckman, eds., Springer, (1995).

....change the relative value of novelty and memory. 2 A Simple Model of Sensorimotor Evolution Our evolutionary system is composed of many agents that could be called organisms, on analogy with biological systems. It has been used in various studies of the evolution of sensory motor functionality [30, 6, 4, 3, 7, 31]. The system consists of many agents that exist together in an environment, in this case a toroidal lattice. The lattice has a real valued field defined on it, E (x) which may be interpreted as an resource field. The resource field is driven by periodically adding resources (from an external ....

....low mutation rates diversity exhibits punctuated equilibrium dynamics, and at very high mutation rates diversity dynamics are random looking fluctuations. Intermediate mutation rates show more complicated fluctuations. Further details about these different diversity dynamics are found elsewhere [4, 3]. Furthermore, total diversity D is dominated by within gene diversity W at low mutation rates, and by between gene diversity at high mutation rates B, as the correlations among the diversity dynamics shows (Figure 3) To further resolve the relationship among the different diversity components, ....

Bedau, M. A. 1995. Three illustrations of artificial life's working hypothesis. In: W. Banzhaf and F. H. Eeckman (Eds.), Evolution and Biocomputation: Computational Models of Evolution. Springer, Berlin, pp. 53--68.

....and micro (individual genomes) levels; ffl explore possible explanations for discrepancies observed between Echo and Holland s theory of CAS. This last topic fulfills Holland s ambition that Echo might be used in the classic hypothesize test revise cycle. We applaud this ambition (see [Bedau, 1995]) and try to apply it here. 2 Holland s Theory of Complex Adaptive Systems Holland s general theory of complex adaptive systems can be summarized in terms of four properties and three mechanisms that Holland thinks all CAS share: ffl Aggregation (Property) This property captures the ....

Bedau, M. A. (1995). Three illustrations of artificial life's working hypothesis. In Banzhaf, W. and Eeckman, F., editors, Evolution and Biocomputation--- Computational Models of Evolution. Springer.

....can smoothly drop to nothing or smoothly rise from nothing. It is obvious enough that evolving systems level of supple adaptation can rise or fall continuously. In fact, there are methods for quantifying various aspects of an evolving system s level of supple adaptation (Bedau and Packard 1992; Bedau 1995), and this enables the dynamics of supple adaptation in artificial and natural systems to be compared directly (Bedau, Snyder, Brown and Packard 1997; Bedau, Snyder, and Packard 1998) Thus, if we view life as supple adaptation, then being alive is a matter of degree. In addition to asking whether ....

....the continuum is the truth underlying the dichotomies which it can be used to define. There is a pragmatic dimension of the issue whether life at bottom is boolean or continuous. If we quantify a system s level of supple adaptation in the way proposed by Bedau and Packard (Bedau and Packard 1992; Bedau 1995; Bedau 1996; Bedau, Snyder and Packard 1998) then one needs a certain amount of data, and so a certain amount of time to gather the data, in order to determine (to within a certain level of statistical confidence) whether a system has a given level of supple adaptation. So, a system exhibiting ....

Bedau, M. A. 1995. Three Illustrations of Artificial Life's Working Hypothesis.

....other macrostates; it might be a pattern of S s behavior, possibly including other macrostates of S. There are also more complicated cases, in which the macrostate is supple or fluid , and the structural definition of the macrostate might be infinitely long. This latter issue is developed in Bedau (1995c) 4 This definition is explicitly restricted to a given macrostate of a given system with a given microdynamic. This is the core or focal notion in a family of related notions of weak emergence, all others of which would be defined by reference to the core notion and would crucially invoke ....

.... agents which demonstrates how these sorts of supple, macro level evolutionary dynamic can emerge implicitly from an explicit microdynamical model (Packard 1989, Bedau and Packard, 1992; Bedau, Ronneburg, and Zwick, 1992; Bedau and Bahm, 1993 and 1994; Bedau 1994; Bedau and Seymour, 1994; Bedau 1995a) What motivates this model is the view that evolving life is typified by a population of agents whose continued existence depends on their sensorimotor functionality, i.e. their success at using local sensory information to direct their actions in such a way that they can find and process ....

[Article contains additional citation context not shown here]

Bedau, M. A. 1995a . "Three Illustrations of Artificial Life's Working Hypothesis." In Banzhaf and Eeckman (1995), pp. 53-68.

....With many models, and especially with unrealistic models, the model s key emergent properties can be observed only by means of certain macro level statistics. Thermodynamic analogues are statistics like pressure and temperature. Some examples of such statistics are diversity of genotypes [2] or species [13] or complexity of taxa (see Ref. 14] and references cited therein) or adaptive evolutionary activity [3, 5] A statistic enables you to see a model s macro scale forest in the face of all the micro scale trees. In addition, these statistics are what allow you to compare the ....

Bedau, M. A. 1995. Three illustrations of artificial life's working hypothesis. In W. Banzhaf and F. Eeckman, eds., Evolution and Biocomputation--- Computational Models of Evolution (pp. 53--68). Berlin: Springer.

No context found.

Bedau, M. A. (1995). Three Illustrations of Artificial Life's Working Hypothesis. In W. Banzhaf and F. Eeckman (Eds.), Evolution and Biocomputation---Computational Models of Evolution (pp. 53--68). Berlin: Springer.

....Comparing evolutionary activity waves within or between systems can show how these evolutionary phenomena vary as a function of time, space, mutation rate, mode of selection, or other factors. In addition, the data displayed in this visualization method can be quantified with various statistics [13, 7, 11, 12], thus enabling evolutionary activity in various artificial and natural systems to be directly compared [11, 12] And since one can argue that the nature of life is intrinsically connected with adaptive evolutionary activity [8, 9] this visualization method can put discussions of the nature of ....

.... behavior of our original Evita run were due to adaptation and which can be attributed to non adaptive factors such as chance (e.g. random genetic drift) or necessity (e.g. the system s underlying architecture) A series of related studies has similarly exploited non adaptive or neutral models [7, 11, 12]. 3 Interpreting Activity Wave Diagrams Graphs of activity distribution functions depict how evolutionary activity of the genotypes (on the y axis) varies as a function of time (on the x axis) The most evident feature of such graphs is the presence of myriad lines or waves (as we will call ....

[Article contains additional citation context not shown here]

Bedau, M. A. 1995. Three illustrations of artificial life's working hypothesis. In W. Banzhaf and F. Eeckman, (Eds.), Evolution and Biocomputation---Computational Models of Evolution (pp. 53--68). Berlin: Springer.

....Comparing evolutionary activity waves within or between systems can show how these evolutionary phenomena vary as a function of time, space, mutation rate, mode of selection, or other factors. In addition, the data displayed in this visualization method can be quantified with various statistics [11, 6, 10], thus enabling evolutionary activity in various artificial and natural systems to be directly compared [10] And since one can argue that the nature of life is intrinsically connected with adaptive evolutionary activity [7, 8] this visualization method can put discussions of the nature of life ....

....In the neutral analogue, by contrast, a genotype s concentration reflects only dumb luck, so no genotype activity waves rise significantly above their peers. This difference between the normal and neutral activity data can be used to quantify the adaptive evolutionary activity in evolving systems [11, 6]. 4 Details in Individual Activity Wave Diagrams In this section we discuss the evolutionary phenomena visible in a handful of individual Evita runs. This is the best way to convey how activity wave diagrams can depict a variety of evolutionary phenomena, such as competitive exclusion and random ....

[Article contains additional citation context not shown here]

Bedau, M. A. 1995. Three illustrations of artificial life's working hypothesis. In W. Banzhaf and F. Eeckman, (Eds.), Evolution and Biocomputation---Computational Models of Evolution (pp. 53--68). Berlin: Springer.

....systems. Evolutionary Activity Statistics Our classification of evolutionary dynamics is based on statistics for quantifying adaptive evolutionary phenomena. These statistics have already been applied to various evolving systems in various ways for various purposes (Bedau Packard 1992; Bedau 1995; Bedau et al. 1997; Bedau Brown 1997) This section describes these statistics with maximal generality and then explains how they are applied here. Our evolutionary activity statistics are computed from data obtained by observing an evolving system. In our view an evolving system consists of a ....

....1. What should be counted as a component, and what counts as the addition or subtraction of a component from the system In most evolving systems components may be identified on a variety of levels. Previous work has studied components on the level of individual alleles (Bedau Packard 1992; Bedau 1995) as well as genotypes (Bedau et al. 1997; Bedau Brown 1997) and taxonomic families (Bedau et al. 1997) Here we study entire genotypes and taxonomic families. The addition or subtraction of a given component consists of the origination or extinction of a given genotype or taxonomic family. It s ....

[Article contains additional citation context not shown here]

Bedau, M. A. 1995. Three illustrations of artificial life's working hypothesis. In Banzhaf, W., and Eeckman, F., eds., Evolution and Biocomputation--- Computational Models of Evolution. Springer.

....evolving systems. We define an innovation as the introduction of a new component into the system. In the case of Evita, the components are entire genotypes. In the case of Bugs, they are also genotypes, though in previous studies, innovations on the level of individual alleles have been measured [7, 4]. For the fossil record, components will be taxonomic families; an innovation is the appearance of a family in the fossil record. 2. How should a given innovation contribute to the evolutionary activity of the system We measure activity contributions by attaching a counter to each component of ....

....their local environment, moving as a function of what they sense, ingesting the resources they find, and reproducing or dying as a function of their internal resource levels. The Bugs model is in a line of models that originated with Packard [26] and has subsequently been evolving in various hands [7, 8, 3, 5, 4, 6, 11] The Bugs model s spatial structure is a grid of sites with periodic boundary conditions, i.e. a toroidal lattice. Besides the agents, all that exists in the world are 50 tiny (3 Theta3 sites) square blocks of resources, which are spread over the lattice of sites and replenished as needed from an ....

Bedau, M. A. 1995. Three illustrations of artificial life's working hypothesis. In W. Banzhaf and F. Eeckman (Eds.), Evolution and Biocomputation--- Computational Models of Evolution (pp. 53--68). Berlin: Springer.

No context found.

Bedau, M.A. 1995. Three Illustrations of Artificial Life's Working Hypothesis. In Evolution and Biocomputation: Computational models of evolution. W. Banzhaf & F.H. Eeckman, Eds. SpringerVerlag, Berlin.

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