|Collections||Selected Papers A-D||Selected Papers E-L||Selected Papers M-Z|
Many may need to be decompressed from their stored format (e.g. z, gz), using WINZIP or a similar utility into .ps form, then printed to a postscript compatible printer. They can often be converted to screen or conventional printer format by using the Ghostscript or Ghostview postscript emulation utilities. PDF files need the free Adobe Acrobat reader. Online abstracts are often included in the collections.
For other papers see Themes , Online Papers , Related Papers and Book Reviews pages .
These links give access to researchers whose work is not available in html format.
Pete Angeline - ALife, EC and GP
Auckland Computer Science - complexity theory
Wolfgang Banzhaf - GP and self organisation in binary strings
Erich Bornberg-Bauer - protein folding landscapes
Rodney Brooks - subsumption robotics
Caltech Avida group - artificial life
CCNR Sussex - evolutionary computing
Chicago University - computer science
Bill Clancey - situated cognition
Carlos Coello Coello - evolutionary multiobjective optimisation
Cog Shop - humanoid robotics
CogPrints Southhampton/JISC - misc
Colorado State - genetic algorithms
Connectionist Research Group - cognitive modelling
Jim Crutchfield - various
Kalyanmoy Deb - Kanpur genetic algorithm lab
Hugo de Garis - artificial neural network brain
Frank Dellaert - robots
Dynamical & Evolutionary Machine Organization (DEMO) - NNs, GAs and Robots
Digital Genetic Research Group - genetic algorithms
ECAL 97 - 4th European Conference on Artificial life
Gerald Edelman - neural Darwinism
Edinburgh Evolutionary Computation Group - various
EEAAX - evolutionary computation
Jeff Elman - connectionism
Essex University - computer science
EvCA Group Santa Fe - evolving cellular automata
EVOALG Group Berlin - evolution and optimisation
David Fogel - evolutionary computation
Walter Fontana - artificial chemistry
Stephanie Forrest - GAs and evolution
Fractals, automates, cellulaires et systèmes dynamiques - fractals & chaos
Bernd Fritzke - self-organizing networks
GARAGe - genetic algorithms
David Goldberg/IlliGAL - genetic algorithms
Robert Goldstone - learning and NNs
GRAL Group - neural nets and learning
Inman Harvey - evolutionary robotics
Dave Hiebeler - alife
Geoffrey E. Hinton's - neural networks
IOWA AI Research Group - NNs and learning
Herbert Jaeger - symbolic dynamics and transient attractors
Yasusi Kanada - randomized CAs
Samual Kaski - self-organizing maps
Joshua D. Knowles - evolutionary multiobjective optimisation
John Koza - genetic programming
Loet Leydesdorff - self-organization
Liverpool Biocomputing Group - cells and GAs
Loughborough Nonlinear & Complex System Group - various
Henrik Hautop Lund - evolving robots
David MacKay - neural networks and protein folding
Bruce MacLennan - synthetic ecology
Bill Mcready - optimization and search
Man-Wai Mak - neural networks
Maryland - chaos
James Meiss - high dimensional maps
Filippo Menczer - latent energy environments
Melanie Mitchell - alife and complexity
Madhavan Mukund - asynchronous automata
Heinz Mühlenbein - mathematica and GAs
Ulrich Nehmzow - robotics
Pedro de Oliveira - various
Jan Paredis - coevolutionary GAs
Domenico Parisi - various
Tony Plate - holographic recurrent networks
Mitchell A. Potter - coevolutionary algorithms
Tom Quick - structural coupling
Tom Ray - alife
Santa Fe Institute - complexity and alife working papers
Udo Seiffert - self-organizing maps
Moshe Sipper - cellular programming
Aaron Sloman - cognition & affect
Didier Somette - self-organization
William Spears - genetic algorithms
Luc Steels - self-organization and language
Sussex University - evolutionary and adaptive systems
Peter M. Todd - evolutionary and adaptive simulation
Marco Tomassini - various
UTSC Neural Networks - self-organization
Jan Vaario - complexity
Dan Ventura - quantum evolutionary computation
Vienna Theoretical Biochemistry Institute - state spaces
Paul Vitányi - computational complexity
Gunter Wagner - evolution
Ronald J. Williams - neural networks
Xin Yao - evolutionary computation
Eckart Zitler - evolutionary multiobjective optimisation
These papers are conceptually interesting archived documents and well worth the effort to download, uncompress (if necessary), print and read. File sizes and types are given for information.
A Developmental Model for the Evolution of Artificial Neural Networks
J. C. Astor & C. Adami (576k pdf)
A model of decentralized growth inspired by developmental biology and the physiology of nervous systems. Each individual artificial neuron is an autonomous unit whose behavior is determined only by the genetic information it harbors and local concentrations of substrates. The power of the artificial chemistry used is demonstrated by analyzing engineered (handwritten) genomes that lead to behaviors known from physiology.
A Report on the 1st International Workshop on Learning Classifier Systems
R.E. Smith (39k ps.gz)
Overview of the concepts behind the LCS methodology with a special emphasis on the cooperative nature of these systems. Weak and Strong cooperative effects are compared and a conclusion that "a Learning Classifier System equals a combination of a GA with cooperativity" is reached.
A Roadmap of Agent Research and Development
Nicholas Jennings, Katia Sycara and Michael Wooldridge (287k pdf)
An overview of research and development activities in the field of autonomous agents and multi-agent systems, identifying key concepts and applications and indicating how they relate to one-another. A range of open issues and future challenges are highlighted.
A Survey of Parallel Genetic Algorithms
Erich Cantú-Paz (279k ps.z)
A categorization of the techniques used to operate multiple connected populations in order to solve a single GA problem. Examples and recent advances are described.
Achieving High-Level Functionality through Complexification
Kenneth O. Stanley and Risto Miikkulainen (144k pdf)
This paper proposes a method for finding high-dimensional solutions incrementally, starting with very small genomes and gradually adding new genes over generations. It further proposes that combining complexification with an indirect genetic encoding, reusing genes, can lead to the discovery of highly complex solutions.
Activity-based Pruning in Developmental Artificial Neural Nets
Alistair Rust, Rod Adams, Stella George, Hamid Bolouri (145k ps.gz)
A brain based model for the growth of 3D neural networks by self-organizing processes and their structure optimisation by removing uncompetitive trees.
Adaptive Perceptual Pattern Recognition by Self-Organizing Neural Networks
Jonathan Marshall (277k ps.gz)
Using both excitatory and inhibitory connections to demonstate a method of obtaining multiple categorisation winners in pattern overlap problems, giving probabilistic 'best fit' outputs.
An Evolutionary Approach to Synthetic Biology: Zen and the Art of Creating Life
Tom Ray (105k ps.z)
Using computers to generate instansiations of new forms of life, exhibiting (but not modelling) those properties commonly required as evidence for the existence of life. The environmental and biological philosophy behind the Tierra system is introduced as a way of thinking about synthetic biology and the evolution of the computer equivalent of multi-cellularity and brains.
An Introduction to Agent Based Modelling and Simulation of Social Processes
Armano Srbljinovic and Ognjen Skunka (233k pdf)
An introduction to the worldview underlying agent-based models, some basic terminology, the properties of such models and their expectations for social-scientific research. Special attention is given to verification.
An Overview of Evolutionary Algorithms in Multiobjective Optimization
Carlos M. Fonseca and Peter J. Fleming (44k ps.gz)
Trade-offs between variables are shown to lead to multiple equivalent solutions, even with as few as two objectives. This look at current techniques considers their sensitivity to objective scaling and the ways in which objectives are combined to create the fitness landscape to be searched.
An Overview of Nonmonotonic Reasoning and Logic Programming
Jack Minker (97k ps.gz)
A look at a formalisation of commonsense reasoning, giving a more intuitive conditional approach than that provided by classical logic and one that can deal with inconsistency and contextual axiom validity. The paper looks also at computational complexity.
An Updated Survey of Evolutionary Multiobjective Optimization Techniques: State of the Art and Future Trends
Carlos A. Coello Coello (44k ps.z)
A look at ways of using Genetic Algorithms to optimize multiple conflicting variables at the same time. The advantages and disadvantages of current techniques are listed along with the problems that remain to be solved, especially where there is no unique solution and the Pareto Front allows multiple equally valid compromise solutions.
Ant Algorithms for Discrete Optimization
Marco Dorigo and Gianni Di Caro (339k pdf)
An overview of recent work on ant algorithms, taking inspiration from the observation of ant colonies foraging behavior. Basic biological findings on real ants are overviewed, and their artificial counterparts and ACO meta-heuristic defined. A number of applications to combinatorial optimization and routing in communications networks are described.
Artificial Chemistries - A Review
Peter Dittrich, Jens Ziegler, and Wolfgang Banzhaf (426k pdf)
This article reviews the growing body of scientific work in artificial chemistry. First, common motivations and fundamental concepts are introduced. Second, current research activities are discussed along three application dimensions: modeling, information processing and optimization. Finally, common phenomena among the different systems are summarized.
Artificial-Life Ecosystems: What are they and what could they become ?
Alan Dorin, Kevin B. Korb & Volker Grimm (285k pdf)
A critique of artificial life simulations interms of how they neglect the energetic and material transformations crucial to real ecosystems, along with a critique of ecosystems simulations which neglect the evolutionary aspects emphasised in ALife. The merger of these two aspects is strongly recommended for both fields.
Artificial Societies and the Social Sciences
J. Stephen Lansing (45k pdf)
An interesting comparison between deterministic and statistical science and a look at the possibilities of ALife techniques bridging the gap between them in the treatment of evolving complex systems, especially in the human-based sciences, comprising highly heterogeneous components.
As Large as Life & Twice as Natural: Bioinformatics & the Artificial Life Paradigm
Paulien Hogeweg (48k ps.z)
A philosophical look at the evolution of lifelike systems by using local interactions without goals, and the emergence of intelligent behavior as a side effect of information use by unconstrained agents. Various ALife examples are compared to CA, NN and GA approaches.
Autopoiesis and a Biology of Intentionality
Francisco J. Varela (59k ps.z)
An attempt to indicate some fundamental or foundational issues of the relation between autopoiesis and perception, taking a perspective from within the organism (a teleological or intentional stance) and relating behaviour to self-maintaining responses to environmental coupling and internal functional requirements.
Autopoiesis and Life
Margaret A. Boden (170k pdf)
This papers examines life as defined by Maturana and Varela as a type of self-organization: autopoiesis in the physical space, and compares it with the concept of metabolism typically included in biological definitions of life. It suggests that if life depends on either autopoiesis or metabolism then strong A-Life is impossible and also sees how the theory of autopoiesis challenges concepts familiar in biology and cognitive science.
Basins of Attraction in Network Dynamics: A Conceptual Framework for Biomolecular Networks
Andrew Wuensche (2406k pdf)
Discrete dynamical network state-space is connected into basins of attraction, ideas of modularity suggesting that the overall network is actually made up of semi-independent sub-network graphs. Each sub-network settles into one of a range attractors according to its current state, which if perturbed can cause the dynamics to jump to alternative attractors. This paper describes how such basins of attraction might provide a conceptual framework for biomolecular networks.
Better Than The Best: The Power of Cooperation
Tad Hogg and Bernardo A. Huberman (179k pdf)
Shows that when agents cooperate in a distributed search problem, they can solve it faster than any agent working in isolation. A quantitative assessment of the value of cooperation for solving constraint satisfaction problems is presented through a series of experiments. Results suggest an alternative methodology to existing techniques for solving constraint satisfaction problems in computer science and distributed artificial intelligence.
Bio-Inspired Computing Tissues: Towards Machines that Evolve, Grow, and Learn
C. Teuscher, D. Mange, A. Stauffer & G. Tempesti (665k pdf)
This paper introduces bio-inspired computing tissues that might constitute a key concept for the implementation of "living" machines, together with the POE model that classifies bio-inspired machines along three axes. Their Embryonics project is described by means of the BioWatch application, characterised by multicellular organization, cellular differentiation, fault-tolerance and self-repair capabilities.
Can Evolution Explain How the Mind Works?
Melanie Mitchell (126k ps)
A critique of Evolutionary Psychology and its connection with Sociobiology. This 'specialised' view of mind is contrasted with the 'generalised' tabula rasa (empty mind) model of Social Science. The relevance of evolution in understanding complex mental abilities is recognised, whilst also recognising current limitations in both biology and psychology that limit scientific rigor in this field.
Can Unrealistic Computer Models Illuminate Theoretical Biology ?
Mark Bedau (87k ps)
Deep questions in biology, regarding emergence and self-organization, open-ended evolution, and diversity growth are shared by many other human and inorganic complex systems. This look at generic and simplified complexity models suggests that qualitative differences still remain that constrain the validity of current attempts to model such novelty creating systems.
Cellular Evolution in a 3D Lattice Artificial Chemistry
Duraid Madina, Naoaki Ono and Takashi Ikegami (625k pdf)
A three-dimensional model of the formation of proto-cell structures is introduced. The results demonstrate the emergence of dynamic three-dimensional cellular structures from a \primordial soup", and a variety of self-maintaining structures may be observed, depending on initial conditions.
Complex Adaptations and the Evolution of Evolvability
Gunter Wagner and Lee Altenberg (299k ps)
A merging of the views of biological and computer science researchers on attempts to achieve complexification. The importance of the genotype-phenotype mapping is emphasised in achieving the modularity needed for successful adaptation.
Complex Systems: Science for the 21st Century
U.S. Department of Energy (2949k pdf)
A beautifully illustrated overview, from a materials science perspective, of the possible benefits of complexity and systems thinking in the fields of collective phenomena, materials by design, functional systems, mastery of nature and new tools. The requirement for interdisciplinary cooperation in the implementation of this new program is stressed.
Complexity and Chaos - State-of-the-Art; Overview of Theoretical Concepts
Robert F. Hadley (830k pdf)
A very comprehensive survey of all the complexity concepts and how they are used by variuos researchers around the world. Four classification criteria distilled from an extensive literature review are then described and used to classify and structure concepts, properties, mechanisms and emerging phenomena. It is one of a series of documents dealing with detailed complexity issues.
Connectionism and Novel Combinations of Skills: Implications for Cognitive Architecture
Robert F. Hadley (186k ps)
A look at the benefits of merging a low level connectionist process viewpoint with a classical architecture comprising high level sub-networks of these modules. This permits the recombination of skills that characterises our mental functioning in the generation of novelty in information flow.
Connectionism : Past, Present and Future
Jordon B. Pollack (54k ps.gz)
A survey of the history of the field, often in relation to AI, discussing its sucesses and failures and giving predictions for where it might lead in the future.
Co-Operation and Self-Organization
Christian Fuchs (670k pdf)
There is a lack of cooperation, self-determination, inclusion and direct democracy in modern society due to its antagonistic centralising structures, whether capitalist or socialist. In order to solve the resultant problems our social systems need re-design in terms of 'participatory' economy, democracy and culture, i.e. power in the system is distributed in such a way that all members and concerned individuals can own, produce, decide and live in the system co-operatively.
Cooperation, Reciprocity and Punishment in Fifteen Small-scale Societies
Joseph Henrich, Robert Boyd, Samuel Bowles, Colin Camerer, Ernst Fehr, Herbert Gintis & Richard McElreath (39k pdf)
A critique of the canonical economics assumption of self-interest based upon extensive multiculture social experiments. Sharing behaviour is found to correlate with the payoffs expected from cooperation in particular societies and the extent of market exchange behaviour in daily lives. Cooperative behaviour was found to be far more common than selfishness suggesting evolutionary advantages based on behavioural changes not included in economic (or biological) models.
Cooperative Coevolution: An Architecture for Evolving Coadapted Subcomponents
Mitchell A. Potter and Kenneth A. De Jong (668k pdf)
Solving increasingly complex problems requires effective techniques for creating interacting modules. One of the major difficulties comes in finding computational extensions to our current evolutionary paradigms that will enable such 'niches' to “emerge” rather than being hand designed. This paper describes one methodology for evolving such inderdependent subcomponents as a collection of cooperating species.
Coordination of Group Behaviour through Stigmergy: Past, Present and Future Work
Adam Russell (142k pdf)
A review of the concept of stigmergy (environmental communication) as it applies to self-organization and agent operations. A view based upon the structural coupling of automomous dynamical systems is presented.
Criticality and Scaling in Evolutionary Ecology
Ricard V. Sole, Suzanna C. Manrubia, Michael Benton, Stuart Kauffman, and Per Bak (1526k ps)
A look at the fractal nature of evolutionary processes and their occurance in ecological systems. The origin of these dynamical features is investigated and self-organized criticality suggested as a common feature, leading to power-law predictions of extinction events and spatiotemporal diversity.
Cybernetics and Second-Order Cybernetics
Francis Heylighen and Cliff Joslyn (74k pdf)
Outline of the historical development of cybernetics in relation to systems (first order) and the observer (second order) followed by a detailed look at the main concepts employed. The field is related to more recent developments in the complexity sciences with an emphasis on its relevance to goal-directed and co-evolutionary epistemology.
Designing Modular Artificial Neural Networks
Egbert Boers, Herman Kuiper, Bart Happel, Ida Sprinkhuizen-Kuyper (37k ps.gz)
Reverse engineering of brain growth processes to obtain a phenotype development method employing L-Systems along with genetic algorithms to evolve efficient architectures.
Discontinuity in Evolution: How Different Levels of Organization Imply Pre-Adaption
Orazio Miglino, Stefano Nolfi, Domenico Parisi (267k ps.z)
A simultaneous Genetic, Neural, Behavioural and Fitness model, showing punctuated equilibria at fitness level by having behavioural equivalence of different neural structures, despite continuous mutation at genetic level.
Discrete Dynamical Networks and their Attractor Basins
Andrew Wuensche (906k ps.z)
An overview of recent work involving basins of attraction, order-complexity-chaos measures and categorization in Cellular Automata and Random Boolean Networks, in one, two and three dimensions.
Elephants Don't Play Chess
Rodney Brooks (484k ps.z)
Introducing the physically grounded subsumption based approach to AI, which relies on levels of modular behaviour rather than symbol manipulation. This demonstrates both emergence in real systems and the irrelevance of high level of computational ability to the demonstration of intelligent seeming behaviours.
Emergence of Differentiation Rules leading to Hierarchy and Diversity
Chikara Furusawa & Kunihiko Kaneko (104k ps.gz)
Looks at the cell attractor types resulting from the interaction of multiple cells, showing that these can self-regulate and include types that cannot exist in isolation.
Emergence and Self-Organisation: A Statement of Similarities and Differences
Tom De Wolf and Tom Holvoet (153k ps.gz)
In this paper a historic overview of the use of each concept as well as a working definition, that is compatible with the historic and current meaning of the concepts, is given. Each definition is explained by supporting it with important characteristics found in the literature. They suggest that emergence and self-organisation each emphasise different properties of a system.
Emergence of Social Organization in the Prisoner's Dilemma: How Context-Preservation and Other Factors Promote Cooperation
Michael D. Cohen, Rick L. Riolo, and Robert Axelrod (404k ps.gz)
Comprehensive experimental data on the effects of various combinations of strategy space, interaction processes and adaptive processes on the emergence and sustainability of cooperation in the Iterated Prisoners Dilemma scenario much used in evolutionary research. Results show that continuity of neighbours is highly significant in maintaining high fitness.
Environmental Puppeteer Revisited: A Connectionist Perspective on Autonomy
Tom Ziemke (234k ps)
A look at freeing robots from imposed control and allowing structural coupling with the environment to develop choice by connectionist self-organization and learning. This involves using the environment as a flexible resource rather than a fixed constraint.
Essays on Darwinism : Ontological Foundations (117k ps.z)
Essays on Darwinism : Organismic Darwinism (193k ps.z)
Essays on Darwinism : Genic Organismic Selection (166k ps.z)
A formulation of Darwinism free from biological interpretation. This allows its application to wider fields including that of Artificial Life. This treatment is a detailed and well reasoned creation of an abstract Dawinism and a critique of some of the misconceptions apparent within the more limited biological focus only on genes as selection mechanisms.
Evolvable Self-Replicating Molecules in an Artificial Chemistry
Tim J. Hutton (283k pdf)
Details of Squirm3, a new artificial environment based on a simple physics and chemistry that supports self-replicating molecules somewhat similar to DNA. The self-replicators emerge spontaneously from a random soup given the right conditions. Interactions between the replicators can result in mutated versions that can out-perform their parents.
Evolving 3D Morphology and Behavior by Competition
Karl Sims (1937k ps)
A look at the techniques used for creating Karl's virtual crreatures and their mapping from genotype to phenotype and behaviour. The system used also include physical parameters to simulate a realistic environment.
Evolving Artificial Neural Networks that Develop in Time
Stefano Nolfi and Domenico Parisi (146k ps.z)
Gradual mapping of genotype into phenotype is examined and this temporal development is shown to have important consequences on NN fitness. It drives early maturation of functional structure and but enforces conservatism, mimicing the similar conservatism seen in early biological development plus its later more innmovative aspects.
Evolving Artificial Neural Networks using the Baldwin Effect
E.J.W. Boers, M.V. Borst, I.G. Sprinkhuizen-Kuyper (30k ps.gz)
An algorithm to add nodes to a modular network to allow improvement of performance and removal of multiple spurious attractors.
Evolving Morphologies of Simulated 3D Organisms Based on Differential Gene Expression
Peter Eggenberger (537k ps.gz)
Using gene regulation to drive self-organising cellular structures in organism development, allowing good scaling behaviour from a small amount of genomic information.
Evolving Neural Networks Through Augmenting Topologies
Kenneth O. Stanley and Risto Miikkulainen (432k pdf)
An overview of Topology and Weight Evolving Neural Network (TWEANN) strategies looking at advantages and problems, together with the outline of a new NeuroEvolution of Augmenting Topologies (NEAT) method that grows a minimal network and is said to outperform fixed topology solutions on a benchmark reinforcement learning task.
Evolving Optimal Populations with XCS Classifier Systems
Tim Kovacs (263k ps.gz)
A look at how the XCS extension of LCSs can achieve optimal rule sets as a result of the switch from payoff prediction strength to accuracy. This approach is shown to also be advantageous in niched environments, where multiple payoff peaks are present.
Evolving the Architecture of a Multi-Part Program in Genetic Programming Using Architecture-Altering Operations
John Koza (157k ps)
This paper describes six genetic programming operations, patterned after the naturally occurring operations of gene duplication and gene deletion and is motivated by Ohno's provocative book Evolution by Means of Gene Duplication. It offers the promise of generating more creative evolutionary models.
Evolution of Random Catalytic Networks
S.M. Fraser & C.M. Reidys (574k ps.gz)
Spontaneous occurance of catalytic cycles in sufficiently complex networks of random sequences which cluster in a small area of system state space.
Evolutionary Connectionism and Mind/Brain Modularity
Raffaele Calabretta and Domenico Parisi (228k pdf)
A critique of the dualist split in cognitive science between modular symbolic computation and homogeneous neural network connectionism. An alternative approach is presented which makes it possible to explore interactions between population evolution and individual learning, along with two examples of evolutionary connectionist simulations that show how modular architectures can emerge in evolving populations of neural networks.
Exploiting Open-Endedness to Solve Problems Through the Search for Novelty
Joel Lehman and Kenneth O. Stanley (333k .pdf)
An interesting look at the possibility of solving problems in genetic algorithms by replacing the objective fitness function by a measure of creativity. It is shown that despite having no explicit evolutionary direction this method can significantly outperform traditional methods. An advantage is that by ignoring the fitness landscape the ruggedness or deception present within this no longer matter.
Exploring Phenotype Space through Neutral Evolution
Martijn Huynen (1.8M ps)
Identifies identical structures arising for multiple different RNA sequences, and shows how these neutral groupings cover adjacent state space yet are sufficiently close to alternative structures to permit access to large numbers of adaptive varients.
From Continuous Dynamics to Symbols
Herbert Jaeger (96k ps.gz)
Looks at ways of categorising continuous systems into symbols and introduces the transient attractor as a mathematical model of state space contraction in conditions of fast bifurcation.
From Simple Associations to Systematic Reasoning
Lokendra Shastri and Venkat Ajjanagadde (309k ps.gz)
SHRUTI is a neurologically plausible model allowing connectionist networks to apply reasoning rules and store long term facts. Dynamic bindings are represented as synchronous firings and rules as sequentially propagating rhythmic activity. Long-term storage uses (statically learnt) 'dynamic pattern matchers' to recognise concurrent associations. No system clock is necessary and node requirements scale linearly with knowledge.
Fungus Eaters Approach to Evolution: A View from Artificial Intelligence
Rolf Pfeifer (59k ps.gz)
A look at emotions in psychology and at AI models of this. Six problems evident from those are outlined and a new approach given that takes them into account. The emergent behaviour of these robots strongly suggests intelligence and emotions to onlookers, yet no such concept is included and emotions thus need not be defined.
Fuzzy Logic vs. Niched Pareto Multiobjective Genetic Algorithm Optimization: Part I. Schaffer's F2 Problem
Brian J. Reardon (371k pdf.gz)
A new multiobjective selection procedure for Genetic Algorithms based on the paradigms of fuzzy logic is introduced, discussed, and compared to the niched Pareto selection procedure. This fuzzy logic approach allows the experimental error or ‘uncertainty’ to be accounted for and used to better refine the optimal parameter range .
Fuzzy Multiple Criteria Decision Making: Recent Developments
Robert Fullér and Christer Carlsson (500k ps.gz)
An excellent review of decision making that discards the classical independence assumption and includes both the mutually exclusive interdependence of interacting criteria and the synergistic mutual support aspects of differing objectives. Also included is the idea that there can be multiple alternative solutions or niches and not just one global optimum. Good examples and bibliography too.
Game Theory, Complexity, and Simplicity Part I : A Tutorial (75k ps)
Game Theory, Complexity, and Simplicity Part II : Problems and Applications (96k ps)
Game Theory, Complexity, and Simplicity Part III : Critique and Prospective (121k ps)
Detailed look at the advantages and limitations of competitive and homogeneous approaches to interaction. The need to extend these studies into more cooperative scenarios, with realistic actors having expertise and limitations is emphasised.
Guiding or Hiding: Explorations into the Effects of Learning on the Rate of Evolution
Giles Mayley (92k ps.gz)
A look at the influence of the Baldwin effect and its opposite, Hiding effect, on evolution speed in learning systems and their sensitivity to learning costs.
Nick Jakobi (95k ps.z)
A biologically inspired encoding scheme for the artificial evolution of neural network robot controllers, where an individual cell divides and moves, in response to protein interactions with an artificial genome, to form an interconnected multicellular 'organism'. The resultant network is then interpreted as a recurrent neural network robot controller. Results are given of preliminary experiments to evolve robot controllers for both corridor following and object avoidance tasks.
How Can We Think the Complex ?
Carlos Gershenson and Francis Heylighen (210k pdf)
A clearly outlined overview of the main features of complex systems and how they differ from classical thinking. The need for adaptive behaviour and self-organization in dealing with essentially unpredictable complex systems is stressed.
How Cellular Models of Urban Systems Work
Paul M. Torrens (648k pdf)
Detailed look at the the problems and advantages of modelling social systems such as cities from the viewpoint of CA and the modifications to the standard CA model that will need to be made in more realistic agent-based simultations.
How Economists Can Get Alife
Leigh Tesfatsion (218k ps.gz)
A summary of Artificial Life and its application to decentralised market economies. A Trade Network Game (TNG) is used to illustrate self-organization in economics.
How Symbiosis Can Guide Evolution
Richard Watson and Jordon Pollack (60k ps.gz)
A look at how the synergy of co-operative interactions allows new niches to be colonised. This symbiotic scaffolding in turn drives genetic evolution in a form orthogonal to the known Baldwin effect.
How to Evolve Rationality & Something Better
Peter Danielson (84k pdf)
Using simulations to show that rational agents outperfom sub-rational ones, and moral agents outperform rational ones in dynamic social interactions. This leads us to question the emphasis on rational choice in the social sciences and ethics, and suggests an evolutionary modelling strategy on which to ground research into moral theory.
Implications of Creation
David Hiebeler (143k ps.gz)
The difficult moral and philosophical issues related to the successful creation of artificial Life are examined. Having a general awareness of these issues is necessary for scientists, whose work does not take place in a moral vacuum but impacts potentially and perhaps devestatingly on our world and human society.
Inspirational Chaos: Executive Coaching and Tolerance of Complexity
Peter J. Webb (87k pdf)
A paper that deals with leadership in a non-Normal world, especially with methods for developing executive wisdom as a strategy for tolerance of complexity, with its associated unpredictability, intractability and short-term solution stability.
Introduction of Structural Dissolution into Langton's Self-Reproducing Loop
Hiroki Sayama (221k ps.gz)
Adding dissolution to the death already incorporated in Langton's self-reproducing CA allow space to be cleared. This may enable potential evolution and complex dynamics in future SRLs.
Introduction to Random Boolean Networks
Carlos Gershenson (379k pdf)
An excellent technical overview of different types of Boolean networks and their properties covering both synchronous and asynchronous types and their various applications. Good references are also provided.
Introduction to the Statistical Theory of Darwinian Evolution
Luca Peliti (98k ps.gz)
A mathematic overview of modern models of co-evolutionary and epistatic genetics, related to the structure of fitness landscapes. Phase transitions and quasispecies are included, taking the treatment beyond conventional population genetics, with analogies drawn also to the relevant physics concepts.
Is Anything Ever New ? : Considering Emergence
Jim Crutchfield (64k ps.z)
A synthesis of tools from dynamical systems, computation and inductive inference is used to outline how we can discover 'novelty in our world, ideas that go beyond our current understanding. Extrinsic emergence (with respect to an external observer) is contrasted with Intrinsic emergence (meaningful in terms of the system itself).
Is There Another New factor in Evolution ?
Inman Harvey (78k ps.gz)
The effect of lifetime learning on uncorrelated tasks in speeding evolution towards quite different goals is examined. It is shown to relate to the spontaneous recovery of perturbed associations when learning unrelated tasks.
Lamarkian Evolution, The Baldwin Effect and Function Optimisation
Darrell Whitley, V.Scott Gordon & Keith Mathias (176k ps)
Comparing Lamarkian and Baldwin forms of evolution with standard GAs in hybrid genetic search shows that the Baldwin method can find global optimums, whilst the others find only local optima.
Landscapes and Molecular Evolution
Peter Schuster (1071k ps)
The importance of neutral networks of equally fit RNA molecules in shape space covering and searches of state space is examined. These 'quasispecies' are shown to allow escape from local optima and efficient search for the global optimum by a combination of adaptive walks and random drift.
Learning from Mistakes
Dante Chialvo and Per Bak (380k ps)
A biologically plausible neural network model using inhibition and no positive reinforcement along with a winner take all selection is shown to self-organize to achieve better learning and expandability than conventional back propagation methods. This synaptic Darwinism requires both negative feedback and extremal dynamics but outperforms more complicated models.
Life and Evolution in Computers
Melanie Mitchell (491k ps.gz)
A review of how far we have come in creating intelligent and alive systems, covering self-reproducing automata, genetic algorithms, decentralised computing and cellular automata that compute non-local information. These steps suggest that living machines are a possibility, although still a distant one.
Macroevolutionary Algorithms: A New Optimization Method on Fitness Landscapes
Jesus Marin and Ricard V. Sole (512k ps.gz)
An interesting alternative to GAs based instead upon ecological models. The MA uses links between species to judge survival and niche colonisation based on long term evolution (a sort of group selection), and is shown to outperform GAs in many areas.
Measures of Statistical Complexity: Why ?
David Feldman and Jim Crutchfield (138k ps.z)
A useful overview of several recent complexity measures and critique of the concept, contrasting entropy and complexity and their relationship to describing and explaining structure.
Metastable Evolutionary Dynamics: Crossing Fitness Barriers or Escaping via Neutral Paths ?
Erik van Nimwegen and James P. Crutchfield (161k ps.gz)
Analysis of the dynamics of mutation. It is shown that timescales for low fitness barrier transitions are orders of magnitude higher than those needed to cross a neutral entropy barrier, making neutral paths in general the only feasible way of reaching a non-local optimum by mutation.
MetaSystems as Constraints on Variation: A Classification and Natural History of Metasystem Transitions
Francis Heylighen (70k pdf)
A new conceptual framework is proposed to situate and integrate the parallel theories of Turchin, Powers, Campbell and Simon. A system is defined as a constraint on variety. This entails a 2x2x2 classification scheme for “higher-order” systems, using the dimensions of constraint, (static) variety, and (dynamic) variation. The scheme distinguishes two classes of metasystems from supersystems and other types of emergent phenomena.
Methods and Techniques of Complex Systems Science: An Overview
Cosma Rohilla Shalizi (808k pdf)
An excellent survey of the mathematical rigor now available within the complexity sciences, biased somewhat towards biomedical applications. Four areas of complex systems study are introduced - patterns, topics, tools and foundations, of which the third is treated here. Extensive references and study hints are included.
Modeling Complexity: The Limits to Prediction
Michael Batty & Paul M. Torrens (390k pdf)
Look at some of the inherent problems in validating complex system models due to the incomplete and partial nature of any possible modelling definition. The key issues are defined and standard closed equilibrium models related to the open, multistate systems typical of urban science.
Modeling Motivations & Emotions as a Basis for Intelligent Behaviour
Dolores Canamelo (1482k ps)
A protolook at creating a artificial baby using ideas expounded by Marvin Minsky in the 'Society of Mid' and Rodney Brooks 'Subsumption Architecture'. This approach also employs aspects of Tomkins 'Affect Theory' to simulate emotional and cognitive development in behavioural selection.
Modular Interdependency in Complex Dynamical Systems
Richard A. Watson (105k pdf)
Hierarchical modularity is a familiar characteristic of a large class of natural dynamical systems. Although modules may be sparsely connected, dynamical properties of modules may be strongly interdependent, and if functional properties depend on dynamical properties then interactions between modules may be critically important and simple evolutionary processes can be inadequate. This paper revews work on compositional evolution and describes its relation to the dynamical properties of hierarchical systems.
Morphodynamic Models of Communication
Peter Bøgh Andersen (656k ps)
An extensive look at the dynamics of form taking in iteration, bifurcation, autopoiesis, catastrophes and symbolisation. The unstable nature of categories with continuous perturbation is emphasised and a 3D cellular automata model used to combine social, psychological and temporal evolution of learning.
John F. Kolen & Jordon B. Pollock (28k ps.gz)
A discussion of the problem of how to implement many-to-many, or multi-associative, mappings within connectionist models. A new model of multiassociative memory is compared to symbolic and neural models. Experimental evidence is given to demonstrate the feasibility of the proposed concept.
Multiobjective Optimization using the Niched Pareto Genetic Algorithm
Jeffrey Horn, Nicholas Nafpliotis & David Goldberg (216k ps.gz)
Combining MultiAttribute Utility Analysis (MAUA) with GAs allows searching of large problem spaces in multidimensional optimisation. Use of Pareto domination selection and niche pressure allows diverse Pareto Optimum solutions to be found.
Natural Niching for Evolving Cooperative Classifiers
Jeffrey Horn and David Goldberg (128k pdf.zip)
An analysis of the robustness of implicit niching (weak cooperation) in an evolving LCS under severe selection pressure. This sort of group fitness forms a basis for the strong cooperation (diversity) needed to solve problems by communicating subpopulations (species) maintained under niching pressure. The results are generally applicable to all types of evolutionary algorithms.
Non-Uniform Cellular Automata: Evolution in Rule Space & Formation of Complex Structures
Moshe Sipper (74k ps.gz)
Taking population of cells with different rules and using genetic algorithm techniques to evolve their populations to produce cooperative structures.
On Constructing a Molecular Computer
Leonard Adleman (140k ps)
A novel approach to using DNA strings to model logic and allowing massive parallelism to compute solutions to NP complete problems. A 'jar' computer a million times more powerful than our best supercomputer
On Information Sharing and the Evolution of Collectives
Michael Lachmann, Guy Sella & Eva Jablonka (86k ps.gz)
The effects of cooperation on fitness levels based on available information is investigated. Advantages both in reduced individual effort and in improved group organization are shown, even where the information acquisition incurrs a cost.
On the Limits of Bottom-Up Computer Simulation: Towards a Nonlinear Modelling Concept
Kurt A. Richardson (277k pdf)
A caution on treating bottom-up simulations as accurate models (called new reductionism) and also on using complexity metaphors uncritically in management science (called soft complexity). Equifinality shows that models are always approximations and should not be applied in isolation. A qualitative framework (called complexity thinking) is needed to encourage both weak and strong explorations of alternative models.
On the State of the Art of POEtic Machines
Christof Teuscher (197k pdf)
Biological systems grow, live, adapt and reproduce, characteristics that are not truly encompassed by any existing computing system. Machines that combine evolutionary mechanisms (Phylogeny), developmental processes (Ontogeny) and learning algorithms (Epigenesis) are called POEtic machines. This paper provides an overview on actual bioinspired machines and stateoftheart models and looks at future directions
Phase Transitions in a Gene Network Model of Morphogenesis
Ricard V. Solé, Isaac Salazar-Ciudad and Jordi Garcia-Fernandez (427k ps.gz)
Models of biological pattern development are presented using both Turing diffusion style processes and inductive cell to cell interaction processes. The studies show previous absract graph network analysis are robust and go on to determine the expected frequencies of stable and unstable spatial patterns for the two models. These indicate that fine tuning is not required, and that self-organizing patterns are a common feature of such networks.
Phase Transitions in Logic Networks
Bruce Sawhill and Stuart Kauffman (194k pdf)
Analysis of randomly connected Kauffman Nets using an annealed approximation and a look at the relevance of the results to complex dynamical systems. Preliminary results are presented on the properties of nets incorporating bias and redundancy.
Plasticity, Evolvability and Modularity in RNA
Lauren W. Ancel and Walter Fontana (2527k pdf)
Investigation of the effects of mutation on the plasticity of RNA folding sequences, showing a reduction in phenotypic diversity due to canalization of phenotype topology into disjoint neutral networks. This has major implications for the ability of neo-Darwinian evolutionary theory to explain innovation and the evolution of new forms and suggests a combinatorics of modular elements at a higher level under environmental diversity may be the only way of generating novelty.
Play Locally, Learn Globally: The Structural Basis of Cooperation
Jung-Kyoo Choi (1488k pdf)
This paper explores the importance of a payoff-based replicator dynamic on two aspects of group structure: interaction in a public goods game and the cultural transmission of behavioral traits. It shows that of the four population structures given by global and local learning and global and local game interaction, local interaction with global learning provides the most favorable environment for the evolution of cooperation.
Principles of the Self-Organizing System
W. Ross Ashby (58M pdf)
An early look at the foundations of organization, the relation between wholes and parts, comparisons between man-made and natural machines, and what constitutes fitness in context. An early understanding of the need for a balance between order and chaos if environmental success and even intelligence is to be an inevitable result of self-organization.
Samuel Bowles and Herbert Gintis (160k pdf)
Prosocial emotions function in providing guides for action that bypass the explicit cognitive optimizing process that lies at the core of the standard behavioral model in economics. Empirical evidence is given suggesting that such emotions play a role in the public goods game. An analytical model and an agent-based simulation shows that reciprocity, shame, and guilt increase the level of cooperation in the group. An explanation of the long term evolutionary success of prosocial emotions, in terms of both the individual and group-level benefits they confer, is offered.
Reasons, Robots and the Extended Mind
Andy Clark (191k pdf)
Two ways of developing the "embodied, embedded" approach are outlined. The first involved a series of tweaks to basic biological modes of adaptive response, the second depicts advanced reason as depending heavily upon a special kind of hybridization having cascades of symbiotic relationships with knowledge-rich artifacts and technologies. The latter does is said to better match our profile of deep biological continuity with deep cognitive discontinuity.
Science of Self-Organization and Adaptivity
Francis Heylighen (104k pdf)
An excellent review of the concepts and principles of the theory of self-organization, containing many examples and clear explanations of its relevance to science and to future modelling of complex human and natural systems.
Selection of Organization at the Social Level: Obstacles and Facilitators of Metasystem Transitions
Francis Heylighen and Donald T. Campbell (105k pdf)
Strengths and weaknesses of the main social control mechanisms are reviewed: mutual monitoring, internalized restraint, legal control and market mechanisms. Competition between individuals and (fuzzily defined) groups at different levels of aggregation very much complicates evolutionary optimization of society. Some suggestions are made for a more effective social organization, but it is noted that the possible path to social integration at the world level will be long and difficult.
Self-Dissimilarity: An Empirical Measure of Complexity
David Wolpert and William Macready (511k ps)
An interesting look at using system inhomogenity at different spatial and temporal scales as a way of classifying complex system. This future technique could enable us to quantify life and compare the different levels of social and technological systems to their physical and biological equivalents.
Selfish Gene Algorithm
Fulvio Corno, Matteo Sonza Reorda, Giovanni Squillero (78k ps.gz)
A look at optimization using statistical gene frequencies rather that phenotye frequencies, to more accurately represent natural selection in the gene centric view. The technique allows fast convergence by positive feedback towards local optima, and implicit co-operation between alleles, without crossover operators.
Social Cohesion and Embeddedness: A Hierarchical Conception of Social Groups
James Moody and Douglas R. White (484 k pdf, Fig5 32 k pdf)
An operationalization of social embeddedness using connectivity concepts than can usefully distinguish between adhesive (hierarchical) and cohesive (heterarchical) networks and additionally deal with large, distance independent, overlapping and nested groupings. This permits correlation between concepts of power and norms in many social settings and is based on graph theoretical concepts.
Strong Reciprocity, Human Cooperation and the Enforcement of Social Norms
Ernst Fehr, Urs Fischbacher & Simon Gächter (143k pdf)
This paper provides evidence challenging the self-interest assumption that dominates the behavioral sciences and much evolutionary thinking. It shows that many people have a tendency to voluntarily cooperate, if treated fairly, and to punish noncooperators. This 'strong reciprocity' cannot be rationalized as an adaptive trait by the leading evolutionary theories of human cooperation, e.g. kin theory, but is a powerful device for the enforcement of social norms.
Symmetry in Evolution
Phillip L. Engle (231k pdf)
A review of the fundamental discrepancies between modern scientific facts, derived from genetics, biochemistry and paleontology, and some of the assumptions behind neo-Darwinism. An alternative. nonlinear and fractal, macroevolutionary explanation, compatible with complex systems theories and individual embryonic development, is presented.
Systems Dynamics and the Lessons of 35 Years
Jay W. Forrester (333k pdf)
A Review of work in social and business modelling aimed at determining the dynamic behaviour of systems and their divergence from expectations. The tendency of people to make incorrect predictions from valid data is emphasised, along with the need to take account of mental data (experience) as well as the more limited written and numerical formalisms often employed.
Systems Dynamics Meets the Press
Donella H. Meadows (30k pdf)
A look at the difficulty of getting systems ideas over to the public and media. It emphasises the need to change the way the public and press perceive reality if we are to improve behaviour, and reduce the harm done by unsystemic thinking.
The Creation of Novelty in Artificial Chemistries
Dominique Gross and Barry McMullin (210k pdf)
They observe that the world surrounding us perpetually creates novelty. The question examines in this article is whether it is possible to build computer models that are similarly creative. The discussion focuses specifically on articial chemistries and contains a very good outline of the philosophy of creativity in various agent scenarios.
The Cybernetic Cut: Progressing from Description to Prescription in Systems Theory
David L. Abel (993k pdf)
A critique of the belief that organization can come either from deterministic or from random systems, along with a claim that evolutionary algorithms are a contradiction in terms. The dichotomy between physical and mental (choice based) approaches is the necessary distinction called the cybernetic cut, an extension of the epistemic cut proposed by others.
The Emergence of Specialization
Hugues Bersini (261k pdf)
This paper discusses various ways, beyond the obvious possibility of unfavoring multi-specialization by paying a high cost, to allow specialists to survive the presence of generalists. It contrasts 'mutualism' with 'division of labour', the latter implying an external user. The parallelism benefits of specialisms are emphasised.
The Essence of Embodiment: A Framework for Understanding and Exploiting Structural Coupling Between System and Environment
Tom Quick, Kerstin Dautenhahn, Chrystopher Nehaniv & Graham Roberts (93k ps.gz)
A framework is presented in which to understand the relational dynamics that exists between organisms and their environments. This co-evolutionary model provides conceptual tools with which to explain the correllation between organism structure and environmental features.
The Evolutionary Origin of Complex Features
Richard E. Lenski, Charles Ofria, Robert T. Pennock & Christoph Adami (422k pdf)
A task examined using digital organisms within Avida. Populations of digital organisms evolved the ability to perform complex logic functions, requiring the coordinated execution of genomic instructions, by building on simpler functions that had evolved earlier, provided that these were also selectively favoured. However, no particular intermediate stage was essential for evolving complex functions.
The Group Development Process seen through the Lens of Complexity Theory
John Campbell, J. David Flynn & James Hay 174k pdf)
An analyis of the benefits of complexity concepts in facilitating group dynamics. A three stage approach is taken: moving the group from its ordered normality to a chaotic state, then a move towards a complexity state (edge-of-chaos), followed by a return to a more creative order.
The History and Status of General Systems Theory
Ludvig von Bertalanffy (986k pdf)
An outline of the earliest 1930's origins of systems ideas leading to cybernetics and the modern computer-based systems approaches. Most of the 'complexity' ideas, including transdisiplinarity, the organismic perspective, open systems, connectivity, nonlinearity, circular causality, emergence, state space and self-organization are echoed in this earlier and sadly now neglected work, but which remains alive under the auspices of the ISSS.
The Inside and Outside Views of Life
George Kampis (37k ps.gz)
Two fundamental forms of description, endo-physics and exo-physics are distinguished, relating to observer detachment from or presence in the reality in question. This epistemological difference is related to the Omniscience (global viewpoint) and Chameleon (local transformability of models) problems along with its relevance to the artificial modelling of life.
The Jigsaw Model: An Explanation for the Evolution of Complex Biochemical Systems & the Origin of Life
John F. McGowan (88k pdf)
An interesting look at systemic genetic coding as a method of mutation that retains functional building blocks, enabling evolution to canalize search space and operate more efficiently. It also suggests a possible function for 'junk' DNA.
The Semantics of Evolution: Trajectories and Trade-offs in Design Space and Niche Space
Aaron Sloman (168k ps)
A look at ways to explore deign space and the contrast betwen environmental niches and alternative designs. Attention is paid to the coevolutionary self-modifying characteristics of designs and the possibility of discontinuous changes and alterations to niches by designs. The possibility of a interdisciplinary mathematical theory of such generalised fitness landscapes in both biological and artificial systems is suggested.
The Second Cybernetics: Deviation-Amplifying Mutual Causal Processes
Margoroh Maruyama (497k pdf)
A classic 1963 paper exploring the often neglected role of positive feedback in biological and human systems in generating inhomogeneity, i.e. structure, from homogeneity (by what is now called the bufferfly effect). The balancing effect of combining both positive and negative feedback paths and the role this plays in evolution is also discussed.
The Systems Perspective: Methods and Models for the Future
Allena Leonard and Stafford Beer (497k pdf)
An introduction to the general systems perspective and a number of tools and models. Whilst seeming ignorant of more modern complex systems ideas (e.g. Alife, GAs, NNs, Attractors) this 1994 paper gives comparative information on employing the techniques of 'Interactive Planning', 'Living Systems Theory', 'Operational Research', 'Socio-Technical Systems', 'Soft Systems Methodology', 'System Dynamics', 'Total Quality Management' and 'The Viable Systems Model'.
Three Ways to Grow Designs: A Comparison of Embryogenies for an Evolutionary Design Problem
Peter Bentley and Sanjeev Kumar (474k pdf)
Exploration of the use of growth processes, or embryogenies, to map genotypes to phenotypes within evolutionary systems. External, explicit and implicit embryogenies are identified and explained and experimental comparisons with a standard GA performed. The results are surprising, with the implicit embryogeny outperforming all other techniques by showing no significant increase in the size of the genotypes or decrease in accuracy of evolution as the scale of the problem is increased.
Through the Labyrinth Evolution Finds a Way: A Silicon Ridge
Inman Harvey and Alan Thompson (323k ps.gz)
Interesting look at hardware evolution using redundant junk DNA, which uses neutral network search and gene switching within a Species Adaptive Genetic Algorithm framework to improve evolutionary potential. This methodology exposes some erroneous ways of thinking about local optima and suggests that the technique used can convert rugged landscapes into effectively smooth ones.
Time Out of Joint: Attractors in Asynchronous Random Boolean Networks
Inman Harvey & Terry Bossomaier (183k ps.gz)
A look at the differences between the cyclic attractors found in synchronously updated networks and the point and loose attractors found in the randomly updated equivalent.
Toward a Definition of Dynamical Hierarchies
Dominique Groß and Tom Lenaerts (158k pdf)
Hierarchical organization is omnipresent in natural systems. This paper attempts a definition of higher level emergence and considers the problems of applying this to current agent based models whose properties fail to show the same higher level interaction features as are evident in real world structures.
Towards Anticipatory Agents
Bertil Ekdahl, Eric Astor and Paul Davidsson (83k ps)
This paper presents a novel approach to the problem of designing autonomous agents that is based on the idea of anticipatory systems. It is argued that systems based on causal reasoning only are too limited to serve as a proper base for designing autonomous agents. An anticipatory agent, on the other hand, will use reasoning from final cause to guide its current actions.
Using Artificial Physics to Control Agents
William M. Spears and Diana F. Gordon (198k ps.gz)
The use of global forces to emulate physical laws and to allow self-assembly, fault tolerance and self-repair to energe in conjunction with local interactions. This procedure is also shown to allow distributed computation to occur and can create geometric self-organizing regularities.
What are Neuro-Fuzzy Classifiers ?
Detlef Nauck and Rudolf Kreuse (214k ps.gz)
A brief overview at combining neural learning of clustering in data to support the development of fuzzy systems. This is contrasted with the reverse process of fuzzy neural networks that use fuzzy logic idea to help create a neural network system.