The University of Southampton
University of Southampton Institutional Repository

Compositional Evolution: Interdisciplinary Investigations in Evolvability, Modularity, and Symbiosis

Compositional Evolution: Interdisciplinary Investigations in Evolvability, Modularity, and Symbiosis
Compositional Evolution: Interdisciplinary Investigations in Evolvability, Modularity, and Symbiosis
Conventionally, evolution by natural selection is almost inseparable from the notion of accumulating successive slight variations. Although it has been suggested that symbiotic mechanisms that combine together existing entities provide an alternative to gradual, or 'accretive', evolutionary change, there has been disagreement about what impact these mechanisms have on our understanding of evolutionary processes. Meanwhile, in artificial evolution methods used in computer science, it has been suggested that the composition of genetic material under sexual recombination may provide adaptation that is not available under mutational variation, but there has been considerable difficulty in demonstrating this formally. Thus far, it has been unclear what types of systems, if any, can be evolved by such 'compositional' mechanisms that cannot be evolved by accretive mechanisms. This dissertation takes an interdisciplinary approach to this question by building abstract computational simulations of accretive and compositional mechanisms. We identify a class of complex systems possessing 'modular interdependency', incorporating highly epistatic but modular substructure. This class typifies characteristics that are pathological for accretive evolution - the corresponding fitness landscape is highly rugged, has many local optima creating broad fitness saddles, and includes 'irreducibly complex' adaptations that cannot be reached by a succession of gradually changing proto-systems. Nonetheless, we provide simulations to show that this class of system is easily evolvable under sexual recombination or a mechanism of 'symbiotic encapsulation'. Our simulations and analytic results help us to understand the fundamental differences in the adaptive capacities of these mechanisms, and the conditions under which they provide an adaptive advantage. These models exemplify how certain kinds of complex systems, considered unevolvable under normal accretive change, are, in principle, easily evolvable under compositional evolution.
Springer
Watson, Richard A.
ce199dfc-d5d4-4edf-bd7b-f9e224c96c75
Yao, Xin
a1c565ef-baac-4165-a9fa-fe89dce883db
al, et
df099e87-31d7-4ccf-a9fa-b92a380537f9
Watson, Richard A.
ce199dfc-d5d4-4edf-bd7b-f9e224c96c75
Yao, Xin
a1c565ef-baac-4165-a9fa-fe89dce883db
al, et
df099e87-31d7-4ccf-a9fa-b92a380537f9

Watson, Richard A., Yao, Xin and al, et (2002) Compositional Evolution: Interdisciplinary Investigations in Evolvability, Modularity, and Symbiosis Brandeis University, MA. USA, Computer Science, Doctoral Thesis .

Record type: Thesis (Doctoral)

Abstract

Conventionally, evolution by natural selection is almost inseparable from the notion of accumulating successive slight variations. Although it has been suggested that symbiotic mechanisms that combine together existing entities provide an alternative to gradual, or 'accretive', evolutionary change, there has been disagreement about what impact these mechanisms have on our understanding of evolutionary processes. Meanwhile, in artificial evolution methods used in computer science, it has been suggested that the composition of genetic material under sexual recombination may provide adaptation that is not available under mutational variation, but there has been considerable difficulty in demonstrating this formally. Thus far, it has been unclear what types of systems, if any, can be evolved by such 'compositional' mechanisms that cannot be evolved by accretive mechanisms. This dissertation takes an interdisciplinary approach to this question by building abstract computational simulations of accretive and compositional mechanisms. We identify a class of complex systems possessing 'modular interdependency', incorporating highly epistatic but modular substructure. This class typifies characteristics that are pathological for accretive evolution - the corresponding fitness landscape is highly rugged, has many local optima creating broad fitness saddles, and includes 'irreducibly complex' adaptations that cannot be reached by a succession of gradually changing proto-systems. Nonetheless, we provide simulations to show that this class of system is easily evolvable under sexual recombination or a mechanism of 'symbiotic encapsulation'. Our simulations and analytic results help us to understand the fundamental differences in the adaptive capacities of these mechanisms, and the conditions under which they provide an adaptive advantage. These models exemplify how certain kinds of complex systems, considered unevolvable under normal accretive change, are, in principle, easily evolvable under compositional evolution.

PDF watson_thesis_2002.pdf - Other
Download (5MB)

More information

Published date: April 2002
Organisations: Agents, Interactions & Complexity

Identifiers

Local EPrints ID: 262006
URI: http://eprints.soton.ac.uk/id/eprint/262006
PURE UUID: 2dc74c6b-7c0a-42fd-9581-ef7088bb814f

Catalogue record

Date deposited: 21 Feb 2006
Last modified: 18 Jul 2017 08:56

Export record

Contributors

Author: Xin Yao
Author: et al

University divisions

Download statistics

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Atom RSS 1.0 RSS 2.0

Contact ePrints Soton: eprints@soton.ac.uk

ePrints Soton supports OAI 2.0 with a base URL of http://eprints.soton.ac.uk/cgi/oai2

This repository has been built using EPrints software, developed at the University of Southampton, but available to everyone to use.

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×