Kauffman, S. A. 1990. Requirements for evolvability in complex systems: orderly dynamics and frozen components. Physica D 42, 135--152.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....The practical benefit for evolutionary computation of understanding and enhancing evolution s adaptive power is obvious, but the theoretical implications of evolvability are also significant. There is general agreement that evolvability is crucial for understanding the orgin of complex adaptations [9, 10, 21, 34, 23] as well as the process of open ended evolution [32, 26] a central open problem in artificial life [7, 8, 5] In addition, Wagner and Altenberg [1, 34] argue that evolvability can unify and explain a host of seemingly related issues, including epistasis, genetic canalization, genetic modularity, ....

....can unify and explain a host of seemingly related issues, including epistasis, genetic canalization, genetic modularity, developmental constraints, developmental and morphological integration, and biological versatility. Studies of evolvability center on understanding how it originates [21, 1, 34] and how it evolves [33, 9, 10, 1, 34] This paper treats the second issue. Evolvability has been criticized on the grounds that it involves group selection and it bestows a future rather than present benefit of individuals, but previous discussions have thoroughly defended evolvability s ....

Kauffman, S. A. 1990. Requirements for evolvability in complex systems: orderly dynamics and frozen components. Physica D 42, 135--152.

....traditional ALife approach, where a large collection of simple state determined automata are combined according to particular network or topological relations. The large collections allow for emergent phenomena at the level of dynamical attractors (a now classical example is the work of Kauffman [22]) In the limitting case, this approaches statistical physics, and the approaches taken are similar, relying on experimentation and statistical descriptions. Small collections of complex agents: This is the traditional AI approach, where a relatively small collection of highly intelligent agents ....

Kauffman, Stuart A: (1990) "Requirement for Evolvability in Complex Systems: Orderly Dynamics and Frozen Components", in: Complexity, Entropy and the Physics of Information, ed. WH Zurek, pp. 151-192, Addison-Wesley, Redwood City 21

....or downright impossible, to tackle. One of these fundamental questions concerns the conditions under which life arises. If one examines our sole example, organic life, the minimal, necessary conditions seem to be: 1) the existence of self replicating entities that are (2) subject to evolution (Kauffman, 1990; Ray, 1992) If these conditions are implemented in some artificial medium, will life arise Though it is probably premature to provide a definite answer at this point, research carried out over the years has lent insight into the issues involved. This paper provides a glimpse into the questions ....

.... Consider, for example, computer programs written in a standard programming language such as FORTRAN or C. A brute force attempt to apply an evolutionary process that randomly reshuffles such programs, would, in most cases, lead nowhere, churning out a plethora of nonfunctional systems (Kauffman, 1990). In fact, it is possible to evolve computer programs by an evolutionary methodology known as genetic programming (Koza, 1992) However, this requires meticulously setting the evolutionary scenario, including the genomic encoding of programs, and the genetic operators applied to them. 16 Journal ....

[Article contains additional citation context not shown here]

Kauffman, S. A. 1990. Requirements for evolvability in complex systems: Orderly dynamics and frozen components. Physica D, vol.

....of chaos in the context of evolution. In particular, it has been hypothesized that when biological systems must perform complex computation in order to survive, the process of evolution under natural selection tends to select such systems near a phase transition from ordered to chaotic behavior [14, 15, 24]. This paper describes a re examination of one study that addressed these questions in the context of cellular automata [24] The results of the original study were interpreted as evidence that an evolutionary process in which cellular automata rules are selected to perform a nontrivial ....

....of models to use in studying how dynamical behavior and computational ability are related. Similar questions have also been addressed in the context of other dynamical systems, including continuousstate dynamical systems such as iterated maps and differential equations [4, 5] Boolean networks [14], and recurrent neural networks [25] Here we will confine our discussion to CA. With this background, we can now rephrase the broad questions presented in Section 1 in the context of CA: ffl What properties must a CA have for nontrivial computation ffl In particular, does a capacity for ....

S. A. Kauffman. Requirements for evolvability in complex systems: Orderly dynamics and frozen components. Physica D, 42:135--152, 1990.

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