 |
||
| Neural Nets:
Electronic automaton, similar in some ways to cellular automata, that offer a highly simplified model of a brain. As such, neural networks are devices for machine learning that are based on associative theories of human cognition. Using various algorithms and weightings of different connections between "neurons," they are set-up to learn how to recognize a pattern such learning a voice, recognizing a visual pattern, learning robotic control, manipulating symbols, making decisions, and so on. Generally, Neural Nets are composed of three layers: input neurons; output neurons; and a layer in-between where information from input to output is processed. Initially the network is loaded with a random program, then the output is measured against a desired output which prompts a adjustment in the "weights" assigned to the connectivities in response to the "error" between the actual and desired output, and this is repeated many times. In this way, the Neural Network learns. In a sense, a Neural Net has to be able to discover its own rules. Changing the rules of interaction between the "neurons" in the network can lead to interesting emergent behavior. In that way, neural nets are another tool for investigating self-organization and emergence. See: Adaptation; Random Boolean Networks Bibliography: Peak and Frame (1994). Stuart Kauffman’s conception of understanding the evolution of complex, adaptive systems based on the fitness traits of an organism (N) and the inputs of one trait (or gene) to another. Then one can observe the fitness landscapes obtained by manipulating the N’s and K’s. Emergent patterns are thus understood in terms of what rules led to them and what implications they have for fitness. See: Adaptation; Fitness Landscape; Self-organization Bibliography: Kauffman (1995); Kauffman and Macready (1995) Technically, any system where the data points coming from the measurement of the values of its variables can be represented as a curvilinear pattern on a coordinate plane, hence, nonlinear for not-a-line. More generally, a system in which small changes can result in large effects, and large changes in small effects. Thus, sensitive dependence on initial conditions (the Butterfly Effect) in chaotic systems illustrates the extreme nonlinearity of these systems. In a nonlinear system the components are interactive, interdependent, and exhibit feedback effects. Complex, adaptive systems are nonlinear systems. See: Linear; Sensitive Dependence on initial conditions. Bibliography: Eoyang (1997); Goldstein (1994); Peak and Frame (1994) One of the defining characteristics of emergent patterns arising from self-organizing processes is their novelty or innovative character. Indeed, that is why such phenomena are termed "emergent" Ñ they introduce new qualities into the system that were not pre-existing in the system. An example are the novel nature of the "dissipative structures" that arise in nonlinear systems at far-from-equilibrium conditions. This novelty is neither expected, predictable, nor deducible from the pre-existing components. Moreover, this novelty is not reducible to the lower level components without loosing its essential characteristics. An issue, therefore, for practitioners working with complex systems, is to determine what system processes (i.e., "anacoluthian") are necessary for the emergence of novelty. That is, novel outcomes demand novel processes that prompt a system to the production of novel structures and practices. See: Anacoluthian; Bifurcation; Emergence; Far-from-equilibrium; Self-organization Bibliography: Goldstein in Sulis and Combs (1996); Kauffman and Macready (1995); Van de Ven & Garud 1994 Stuart Kauffman’s term for the way the internal dynamics of a system generates order spontaneously under the right conditions. This order is "for free" in that it does not need to be imposed or imported from outside the system. It is Kauffman’s conjecture that natural selection during the course of evolution takes place on already self-organized order. An implication is that particular biological adaptations or forms may result from constraints on possible designs due to the inherent mathematical dynamics of a system. In terms of organizations, it may be case that spontaneously emerging structures provide a crucial understanding of such system’s functioning. See: Emergence; Self-organization Bibliography: Kauffman (1995) |
||
Return to
Top | List of Terms |