 |  |
| |
 |
lingue |
|
<!DOCTYPE HTML PUBLIC '-//W3C//DTD HTML 4.0 Transitional//EN'><!-- saved from url=(0182)https://lw8fd.law8.hotmail.msn.com/cgi-bin/getmsg?disk=216.33.240.66_d745&login=espositogianluca&f=33792&curmbox=ACTIVE&_lang=IT&msg=MSG954797845.10&start=1598166&len=58549&mimepart=5 -->
Artificial Life
A new approach to understanding, building and managing complex systems
Outline
INTRODUCTION
Objectives
Selection of themes
Using the Bibliography
THEMES
Artificial Life: Goals and Methods
Artificial life as a technique for investigating complex system dynamics
Evolutionary computing as a basic tool in Artificial Life Research
'Evolving intelligence: Evolutionary Computing and Artificial Neural Networks combined
Collective behavior: emergent dynamics in populations of interacting agents
Artificial Life and the evolution of complexity
A-Life modelling in specific domains
Artificial Life as a tool in theoretical biology
Rats and robots: Artificial Life as a tool for psychological research
Artificial Life as a tool in the social sciences: simulating cooperation and conflict
Artificial Life as a technology: applications
WEB RESOURCES ON ARTIFICIAL LIFE
Complex systems are hard to build - and difficult to maintain. No engineer is capable of building machines which rival the flexibility and the robustness of even the simplest animals. And the first thing one notes when faced with complex economic, social, political and technological systems is their apparent unpredictability. The extreme difficulty of understanding, building and maintaining complex systems presents a stimulating theoretical challenge for scientists while creating continual practical difficulties for engineers, managers and politicians.
One way of trying to understand the working of complex systems is to look at biology. Darwinian natural selection creates complex self-organized systems, apparently out of nothing. In recent years mechanisms similar to Darwinian evolution have been discovered or hypothesized in a broad set of domains ranging from the immunology to neuro-science, from economics, sociology and linguistics to cosmology. The new science of 'Artificial Life' is based on the basic hypothesis that in each of these cases there is a common set of underlying principles at work. Knowing these principles and simulating their working on computers should make it possible to achieve a better understanding of the functioning of specific kinds of complex system. This in turn should provide guidance into how best to manage such systems. In some cases the knowledge attained is already helping engineers to design machines which are more flexible and robust than those designed by traditional methods.
The tools and concepts used by A-Life researchers have provided useful scientific insights in fields as far apart as theoretical biology, cognitive psychology, economics and political science. At a more practical level A-life techniques have been successfully applied to a broad range of engineering problems which go from building robots to the optimization of airline schedules, from 'drug discovery' to the design of electronic circuitry.
The aim of this seminar is to provide an overview of the new discipline for non-specialists. Given the broad interdisciplinary span of A-life research the course will be of interest to students from a broad range of disciplines including computer science, engineering, biology, psychology, economics, political science and philosophy.
The subject matter treated will include: universal features of complex systems, biological foundations and the Darwinian paradigm, concepts and techniques in A-life modeling, the application of A-Life techniques in 'evolutionary robotics' and cognitive psychology, A-life models of economic, social and political cooperation and conflict and engineering applications of A-Life. The final module of the course will look into some of the philosophical implications of A-Life research with particular reference to the problem of 'the origins of complexity'.
The themes suggested for student papers are divided into two groups:
Artificial Life: goals and methods
A-Life modelling in specific domains
Themes in the first group are intended to cover broad areas of A-Life providing an overview of basic issues and techniques. The more 'philosophical' themes are best suited to students with broad interdisciplinary interests ranging beyond the traditional concerns of Computer Science. The themes covering specific techniques are more technical and will be of most interest to students interested in practical applications of A-Life techniques. The themes in the second group ('A-Life modelling in specific domains') may be treated either with 'overview papers' providing surveys of work in a specific field or with 'monographic' studies going into depth on a particular problem. In this latter case it may be possible for students to perform a limited amount of experimental work, incorporating the results in their paper.
The bibliography provided in the 'Goals and Methods' section is of general interest and may be usefully referred to by all students. The bibliography referring to specific domains will mainly be of interest to students addressing themes in that domain.
The bibliography is NOT a required reading list. The intention is to provide students with a broad initial choice of possible reading materials. Students may use this initial material as a starting point for further reading.
Students are also encouraged to make use of the Web. Click here for a guide to Artificial Life on the Web
Artificial Life: goals and methods
Artificial Life as a technique for investigating complex system dynamics
Goal:
To provide a broad, non-technical, overview of the goals and methods of Artificial Life. This paper is suitable for students with broad inter-disciplinary interests
Background, techniques and issues:
What is a complex system? Examples: the brain, the genome, market economies, eco-systems, telecommunications networks. Understanding, building and managing complex systems.
Characteristics of complex systems: : determinism, chaos and unpredictability.
How can complex systems be understood/designed/managed?
The Darwinian paradigm. Evolution via natural selection. Applications of the Darwinian paradigm outside the traditional domain: the brain, the immune system, political conflict and cooperation, the economy, cosmology.
Computer simulation as a tool for investigating the evolution of complex systems
A-Life vs. traditional modelling: bottom-up vs. top-down, explicit representation of mechanism, the emergence of complexity from essentially simple dynamics
Bibliography:
Bak P., 1996, How Nature Works, Springer Verlag (Popular introduction to the concept of self-organized criticality)
Kauffman S., 1993, The Origins of Order, O.U.P. (General background - a very difficult but extremely stimulating read. Possibly the best book ever written on the subject)
Kauffman S., 1995, At Home in the Universe, The Search for the Laws of Complexity, Penguin Books (a popular rending of the previous reference. Extremely readable)
Kelly K., 1994, Out of Control, The New Biology of Machines, Social Systems and the Economic World. (Popular book. Not always reliable scientifically but gives an excellent 'philopsophical' introduction to the subject.)
Langton C.G. (ed), 1992, Artificial Life III, Addison Wesley (Collection of key papers)
Langton C.G., Taylor C., Farmer J.D., Rassmussen S. (ed)., 1992,Artificial Life II, Addison Wesley (Collection of key papers)
Langton C.G., Ed., 1989,. Artificial Life I,
Smolin L., 1997, The Life of the Cosmos, O.U.P. (Applies evolutionary ideas to cosmology - highly controversial - may not be true but stimulating)
Evolutionary computing as a basic tool in Artificial Life research
Goal:
To provide an overview of basic techniques in 'evolutionary computing' and of the ways in which these techniques have been applied in Artificial Life Research.
Background, techniques and issues:
Basic characteristics of the Darwinian paradigm. Genotype and phenotype. Random variation in the Genotype. Natural selection. Differential reproduction: the 'survival of the fittest'. Sexual and asexual reproduction. Genetic drift.
Genetics: the genome, DNA. Meiosis and Mitosis. Transcription and translation. Mutation. Insertion, Deletion.
Formal models: Fitness landscapes. Hill-climbing and the problem of local optima.
'Artificial evolution'. General goals. Problem representation. Genetic operators. Fitness functions.
Specific techniques: Genetic programming, Evolutionary Strategies, Genetic Algorithms. Choice of parameters. Optimal rates of mutation.
Applications of Evolutionary Computing to A-Life research: applications in theoretical biology, psychology, economics, political science
Bibliography:
Fogel, D.B., 1995, Evolutionary Computation, Towards a New Philosophy of Machine Intelligence, IEEE Press (Discusses the impact of evolutionary computing on Artificial Intelligence)
Holland, J.H.,1975,. Adaptation in Natural and Artificial systems,
Koza, J.R., 1991, Genetic Evolution and Co-evolution of Computer Programs, in Artificial Life II, op. cit. (Evolutionary computing in an A-Life setting)
Koza, J.R., 1992, Genetic Programming, On the Programming of Computers by Means of Natural Selection ;MIT Press (Technical introduction to genetic programming)
Mitchell, M. 1996, An Introduction to Genetic Algorithms, MIT Press (General, technical introduction to Genetic Algorithms)
Smith, J.M. 1998, Evolutionary Genetics, OUP (A basic textbook on evolutionary genetics, useful background material)
(For this theme students may also refer to the bibliography for Combining ANNs with Evolutionary Computing
'Evolving' intelligence: Evolutionary Computing and Artificial Neural Networks combined
Goal:
To provide an overview of how Artificial Neural Networks have been combined with Evolutionary Computing to simulate the evolution of 'intelligent systems'
Background, techniques and issues:
The basic concept of an Artificial Neural Network: artificial neurons, synapses, activation functions. Learning algorithms. Supervised and un-supervised learning. Artificial Neural Networks as universal computers.
Varieties of Artificial Neural Network: Perceptrons. Hopfield networks. Kohonen networks. The Back-proation algorithm.
Evolutionary Computing and Artificial Neural Networks. Issues of representation. Genotypes and phenotypes. ANNs as 'weight matrices'. 'Growing networks' .
Evolving ANNs in A-life research: applications in psychology, economics and political science.
Bibliography:
Artificial Neural Networks
Crick F.1989. The Recent Excitement about neural networks, Nature, 337: 129-l32 (Proposes a critical vision of the biological/psychological of Artificial Neural Networks )
Hopfield J.P., 1982, Neural Networks and physical systems with emergent collective computational abilities, Proceedings of the National Academy of Sciences 79: 2554-2558 reprinted in Neurocomputing op. cit. (A fundamental paper which contributed to making ANNs scientifically 'respectable')
Kohonen T., 1984, Self-Organization and Associative Memory, Springer Verlag (proposes a novel model for unsupervised learning by ANNs - has been used in many applications though not - as far as I know - in A-life research)
Kohonen T., 1988, The Neural Phonetic Typewriter, Computer, 21, 11-22 reprinted in Neurocomputing 2, op.cit. (a very clever application of ANNs to speech perception)
Linsker R., 1988, Self Organization in a Perceptual Network, Computer 21:105-l17 reprinted in Neurocomputing 2, op. cit. (a brilliant application of ANN modelling to the modelling of visual perception in the mammal brain - should be read together with Hubel and Wiesel - see the bibliography for the application of ANNs to psychology)
Minsky M. & Papert S., 1988, Perceptrons, MIT Press (a fundamental though very difficult text, discussing the basic mathematics of ANNs and their inherent limitations)
Rumelhart, D.E., . McLelland, J.L. and the PDP Research Group, 1987, Parallel Distributed Processing, Explorations in the Microstructure of Cognition, MIT Press. (The classic work on Artificial Neural Networks - includes extremely useful 'tutorial material' as well as software)
Rumelhart,
D.E., Hinton, G.E., & Williams, R.J. ,1986,. Learning Internal
Representations by Error Proation in Parallel Distributed Processing eds.
D.E. Rumelhart, J.L. McClelland and the PDP Research Group I: 318-364,
Combining ANNs with Evolutionary Computing
Littman,
M., 1996. Simulations combining Evolution and Learning in Adaptive Individuals
in Evolving Populations, 465-477,
Miglino O., Nafasi K. & Taylor C. ,1995, Selection for wandering behavior in a small robot, Artificial Life, 1995 2-l (reports experiments in which evolutionary techniques are used to produce control networks for robots displaying animal-like exploratory behavior)
Miglino, O., Denaro, D., Tascini, E., Parisi, D. ,1998,. Detour behavior in evolving robots. ,Proceedings of First European Workshop on Evolutionary Robotics (evolves robots displaying the ability to circumvent an obstacle. This kind of 'detour behavior' is classically seen as symptomatic of animal intelligence)
Walker R. & Miglino O., 1999,Simulating exploratory behavior in evolving Artificial Neural Networks, Proceedings of GECCO 1999, Morgan Kauffman, 1999 (A follow-up of the work reported in the preceding paper)
Collective behavior: emergent dynamics in populations of interacting agents
Goals:
To see how A-life models can be used to investigate the collective behavior of populations of interacting agents (e.g. ants in an ant colony, routers in a telecommunications network, cars on city streets, companies in an economy) and ways in which these may lead to the emergence of complex patterns of behavior
Background, techniques and issues:
Ecological models
Populations of organisms competing for a shared resource - the emergence of specialization
Predators and prey.
'Superorganisms' - analogies with the behavior of 'social insects'. Ant Colony Optimization
The evolution of communication and language
Bibliography:
Chialvo D.R., Millonas M.M., How Swarms build cognitive maps, 1995, SFI Working Paper, 95-03-033 (shows how a population of simple agents can generate a map of its external environment - in the same way as a complex agent with a 'brain')
Collins R.J., Jefferson D.R., Antfarm, towards simulated evolution, in Artificial Life II, op. cit. (A-Life model of the evolution of an ant colony)
Dorigo M., Maniezzo V. & Colorni A, 1996, The Ant system: optimization by a colony of cooperating agents, IEEE transactions on systems man and cybernetics op. cit. (Dorigo's work on ant-colony optimization has been a key inspiration for many workers in this field)
Lund H.H. & Mayoh B.H., 1995, Specialization under social conditions in Shared Environments, in Moran F., Moreno A., Merelo J.J., & Chacon (eds), Advances in Artificial Life: Proceedings of the Third European Conference on Artificial Life, Springer Verlag (shows how competition for shared resources can lead to specialization within a population)
Mataric M.J., Designing and Understanding adaptive group behavior, Adaptive Behavior 4 (1) (Use of A-life models to understand group adaptation and behavior)
Steels L., Synthesizing the origins of language and meaning using co-evolution, self-organization and level formation in Hurford J. (ed), The Evolution of Human Language, Edinburgh University Press, Edinburgh (one of a large number of papers in which Steels and his group simulate the evolution of syntax and semantics in a population of cooperating agents)
Artificial life and the evolution of complexity
Goal:
To examine general issues concerning the evolution of complex systems and to see the contribution which A-Life models can make to achieving improved understanding of the necessary and sufficient conditions for such evolution
Background, techniques and issues:
Formal definitions of complexity. The concept of algorithmic complexity. The concept of NP-completeness.
'Order for free'. Self-Organized Criticality.
Complexity in biological evolution. Measuring biological complexity. The size of the genome in different species. Redundancy in the genome. 'Junk DNA'. Does biological complexity increase over time? Co-evolution as a source of complexity.
Modelling the evolution of complexity with A-Life models. Variable length genomes. Coevolutionary models. Modelling the role of the environment. 'Niche selection'. Modelling the emergence of redundancy. Gene and chromosome duplication. Growing networks. Modelling the emergence of modularity.
Extending model results to other domains. The evolution of complexity in ecological, economic, social and political systems.
Redefining traditional concepts of what constitutes a living system.
Bibliography
Garey M.R. & Johnson D.S. , Computers and Intractability, 1979, W.H. Freeman & Company (a classic introduction to the concept of complexity and NP-completeness.)
Chaitin, G.J. 1975. Randomness and Mathematical Proof, Scientific American, 232:5 47-52 (a readable and lucid introduction to the concept of algorithmic complexity - by one of the inventors of the concept)
Kauffman S., 1993, The Origins of Order, O.U.P. (General background - a very difficult but extremely stimulating read. Possibly the best book ever written on the subject)
Kauffman S., 1995, At Home in the Universe, The Search for the Laws of Complexity, Penguin Books (a popular rending of the previous reference. Extremely readable)
Judd, J.S.,1990, Neural Network Design and the Evolution of Complexity, MIT Press (a relatively unknown but important work showing that certain problems in Neural Network learning are NP-complete)
Gould, S.J.,1996, Full House, Three Rivers Press (a popular book arguing that there is in fact no general tendency for biological systems to become more complex over time - the most successful biological systems on the the et are of course bacteria.)
Eigen, M. & Schuster P., 1978., The Hypercycle Part B: ``The Abstract Hypercycle'', Naturwissenschaften, 65: 7-41. (an extremely demanding, highly theoretical yet very useful introduction to the first stages in the evolution of life)
A-Life modeling in specific domains
Artificial Life as a tool in theoretical biology
Goal:
To investigate how techniques from Artificial Life can be applied to the investigation of issues in theoretical biology. Papers could provide a broad overview of the application of A-Life techniques to biological issues. Alternatively a student could make an in-depth investigation of a specific issue in current biological research, perhaps including results from small-scale simulations.
Background, techniques and issues:
Much work in theoretical biology is concerned with investigation of the general dynamics of evolutionary processes. Some of this work (e.g. work on the evolution of complexity or the evolution of cooperation) has implications for fields outside biology. Examples of issues which have been investigated using A-Life techniques include:
'Punctuated' vs. 'Gradualistic' models of biological evolution
The relationship between genotype and phenotype - 'development'
The evolution of sex
The role of 'genetic drift' (random changes in the genome unrelated to selective pressures)
Group vs. individualistic selection: the evolution of cooperation
Co-evolution (e.g. of predator and prey)
Redudancy in the genome and the role of 'junk DNA'
Gradual and catastrophic environmental change as agents of biological evolution
Bibliography:
Background material on biological evolution
Dawkins R., 1982, The Extended Phenotype, O.U.P. (a biologist's attempt to apply evolutionary ideas outside the domain of biology - extremely well written)
Gould, S. J Lewontin, R. ,1979,. The Spandrels of San Marco and the Panglossian Paradigm: A Critique of the Adaptationist Programme. ,Proceedings of the Royal Society of London, Series B 205(1161), 581-598 (Highly effective critique of the standard, pan-adaptationist view of Darwinian evolution)
Kimura M., 1983, The Neutral Theory of Molecular Evolution, Cambridge University Press (very important but extremely difficult book introducing the concept of 'neutral' evolution)
Smith J.M. & Szathmary E., 1995, The Major Transitions in Evolution, W.H. Freeman (a technical account of 'macro-evolution' , demanding but worthwhile)
Smith J.M. , 1958, The Theory of Evolution, Penguin (a basic introductory text - non technical but essential reading)
Smith J.M., (1998), Evolutionary Genetics, O.U.P. (a basic, clearly written textbook on evolutionary genetics - useful background material)
Weiner
J., 1994, The Beak of the Finch, Knopf (a wonderful account of the empirical
testing of
Wright S.,1931,. Evolution in Mendelian Populations, Genetics, 16:97.(introduces the fundamental concept of a fitness landscape)
Applications of A-Life models in theoretical biology
Bagley R.J. & Farmer J.D., 1991, Spontaneous Emergence of a metabolism, in Artificial Life II, op. cit. (addresses the question of the emergence of simple cellular metabolism)
Kauffman S.A, and Johnson S., 1991, Coevolution to the Edge of Chaos, Coupled fitness landscapes, Poised States and Co-evolutionary avalanches, in Artificial Life II, op. cit. (introduces the concept of 'evolution to the edge of chaos' - perhaps A-Life's most important proposed contribution to theoretical biology. A classic paper)
Miglino O., Nolfi S. & Parisi D., 1995, Discontinuity in evolution: how different levels of organization imply pre-adaptation, in R.K. Belew and M. Mitchell (eds.) Plastic Individuals in Evolving Populations, Addison Wesley (models a punctuationalist view of evolution)
Ray T.S., 1991, An approach to the synthesis of life, in Artificial Life II, op . cit. (Introduces 'Tierra' - a classic A-Life model)
Rats and robots: Artificial Life as a tool for psychological research
Goals:
To show how models using Artificial Life techniques can provide useful insight into the evolution and functioning of animal and human intelligence. Papers dedicated to this theme could provide a broad overview of current work. Alternatively students could provide an in-depth account of a single theme including (optionally) the results of small-scale simulations.
Background, techniques and issues:
Over the last twenty years much research in psychology has taken a 'cognitivist' approach in which human/animal brains are modelled as information processing systems performing symbolic manipulations on a formalized representation of the outside world. A-life modellers have challenged this view designing models which match the observed performance of animals without resort to 'symbolic' information processing.
Theoretical issues raised by A-life research include:
The importance of 'learning' in adaptation
The role of the body in biological 'intelligence'
The role of the environment as a 'store of information'
The need for/nature of 'representations' in biological organisms
The evolution of a 'modular' intelligence
The choice of tools to be used in this kind of investigation is the subject of keen debate with different researchers concentrating their attention on:
Software models of artificial organisms ('animats')
Hardware models of animal intelligence (robots)
'Evolutionary robotics' (combining Evolutionary Computation and hardware models)
Experimental work has concentrated on the simulation of 'low-level' animal-like intelligence. Successful models have included:
Models of animal gait (the way an animal walks)
Models of animal navigation and obstacle avoidance (e.g. in various forms of maze)
Models of mating behavior
Models of 'flight' and 'pursuit'
Bibliography:
Background material on the brain
Hubel D.H., 1988, Eye, Brain and Vision, Scientific American Books (a classical work describing experiments to elucidate the workings of the vision system in mammals. For this work Hubel and Wiesel won the Nobel prize)
Zeki S., 1993, A Vision of the Brain, Blackwell (a general description of the workings of the brain with particular emphasis on vision)
Background material on animal psychology
General works
Gallistel.C.R.,1990, The Organization of Learning, MIT Press (a classical text-book)
Gould J.L. & Gould, C.G., 1994, The Animal Mind, Scientific American Library (an extremely well-written book for the general reader showing that animal intelligence is, in fact, far more complex than generally recognized)
Experimental and theoretical work referred to in A-life models
Bennett, A. T. D., 1996. Do animals have cognitive maps?. Journal of Experimental Biology, 199 (1), 219-224. (discusses the existence of cognitive maps in animals - arriving at sceptical conclusions)
Hebb, D. 1949, The Organization of Behavior, New York, Wiley, extract in Neurocomputing op. cit. (Develops the basic model of memory and representation on which virtually all ANN work is based)
Regolin, L., Vallortigara, G., Zanforlin, M. ,1994, Object and Spatial Representations in Detour Problems by Chicks. Animal Behavior, 48:1-5.
Roberts D.L., & Good M. ,1999,. Hippocampal lesions disrupt navigation based on cognitive maps but not heading vectors, John M. Pearce, Amanda, Nature, 396, 75-77 (reports neurological experiments supporting the existence of cognitive maps)
Tolman, E. C. ,1948,. Cognitive maps in rats and men. Psychological Review, 36, 13-24. (reports experiments alleged to demonstrate the existence of 'cognitive maps' in rats)
Empirical work in A-Life
Burgess N., Donnett J.G., O'Keefe, J., 1997 Robotic and Neuronal Simulation of Hippocampal navigation, AISB Workshop on 'Spatial reasoning in Mobile Robots and Animals, Technical Report n. UMCS-75-4-l, University of Manchester. (This paper explicitly es results from neurological research with A-Life simulations of animal navigation)
Cliff D. & Miller G.F.,1996, Co-evolution of Pursuit and Evasion II: Simulation methods and Results, in From animals to animats 4: Proceedings of the Fourth International Conference pm Simulation of Adaptive Behavior (SAB96), Maes P, Meyer J-A, Pollack J. & Wilson S.W. (eds), MIT Press. (This paper shows how A-life methods can be used to simulate the coevolution of behavior in predators and prey)
Lund H. H., Webb B., and Hallam, J. A ,1997,. Robot Attracted to the Cricket Species Gryllus bimaculatus. In P. Husbands and I. Harvey, editors, Proceedings of Fourth European Conference on Artificial Life, es 246--255, Cambridge, MA,. MIT Press, Bradford Books. (reports the evolution of a robot capable of simulating the mating behavior of a cricket)
Miglino O., Nafasi K. & Taylor C. ,1995, Selection for wandering behavior in a small robot, Artificial Life, 1995 2-l (evolving robots exhibiting animal like exploratory behavior)
Miglino, O., Denaro, D., Tascini, E., Parisi, D. ,1998,. Detour behavior in evolving robots. ,Proceedings of First European Workshop on Evolutionary Robotics (reports the evolution of robots capable of exhibiting 'detour behavior' - often regarded as symptomatic of animal intelligence)
Walker R. & Miglino O., 1999,Simulating exploratory behavior in evolving Artificial Neural Networks, Proceedings of GECCO 1999, Morgan Kauffman, 1999 (builds on the work reported in the previous reference)
A-Life and cognitive theory
Beer, R.D.,1990, Intelligence as Adaptive Behavior, Academic Press (proposes a bottom-up view of intelligence - with specific reference to the simulation of insect 'gait'.)
Braitenberg, V. 1984, Vehicles, Experiments in Synthetic Psychology, MIT Press (a classic yet essentially simple book, developing 'thought experiments' which A-Life researchers were later to put into practice)
Brooks, R.A. ,1991, Intelligence without representation. Artificial Intelligence, 47 (1-3), 139-l59. (A classical paper proposing the view that it is possible to evolve animal-like intelligence without resort to the kind of representations proposed by cognitivist models)
Brooks, R.A.,1990, Elephants don't play chess, in P. Maes (ed.) Designing autonomous agents, MIT Press (emphasizes the importance of 'low-level, non-symbolic intelligence)
Clark, A.,1997, Being There, MIT Press (A beautiful book on the contribution of A-Life to psychology - essential reading)
Lettvin, J.Y., Maturana, H.R. McCullough, W.S., Pitts, W.H., 1959, What the frog's eye tell's the frog's brain, Proceedings of the Instute of Radio Engineersm 47: 1940-51. Reprinted in Anderson, Pellionisz and Rosenfeld, eds, 1990,Neurocomputing 2, MIT (Perhaps the first paper developing a neural network like model of animal cognition - a classic)
Rumelhart, D.E., McLelland, J.L. and the PDP Research Group, 1987, Parallel Distributed Processing, Explorations in the Microstructure of Cognition, MIT Press (this essential and readable book is an early attempt to apply ANN models to research issues in psychology)
Technical issues in the use of A-life models for psychological research
Miglino O., Lund H. H., Nolfi S.,1996, Evolving Mobile Robots in Simulated and Real Environments, Artificial Life, 2(4): 417-434 (emphasizes the difference between real and simulated environments)
Mondada F., Franzi E. Ienne P. ,1993, Mobile Robot miniaturisation. A tool for investigation in control algorithms. In Proceedings of the Third International Symposium on Experimental Robotics, Kyoto, Japan (Introduces the 'Khepera' robot, often used in A-life experiments)
Nolfi S. & Parisi, D., 1993, Self Selection of input stimuli for improving performance, in C.A. Bekey, (ed.) Neural Networks and Robotics, Kluwer (shows how autonomous agents select their own inputs therefore simplifying the task of learning and adaptation)
Nolfi, S., Floreano D., Miglino O., Mondada F., 1994, How to evolve autonomous robots: different approaches in evolutionary robotics. In Brooks R.A. & Maes P. (eds), Artificial Life IV- Proceedings of the Fourth International Workshop on the Synthesis and Simulation of Living Systems, pp. 190-l97, Cambridge MA, MIT Press (An overview of different approaches to evolutionary robotics)
Artificial Life as a tool in the social sciences: simulating cooperation and conflict
Goal:
To show how techniques from Artificial Life can be applied to the study of cooperation and conflict in social, economic and political systems. Although this theme is primarily of concern to social scientists it may also be of interest to theoretical biologists interested in the emergence of 'altruism'. The theme will be of particular interest to students with a background in the Theory of Games.
Background, techniques and issues:
Many animal and human societies show a degree of altruism - individuals within the society sacrifice their individual interests for the interests of the group. This phenomenon is difficult to explain in purely Darwinian terms. At first sight it would appear that a society of 'cooperators' is prone to 'invasion' by 'parasites' who benefit from the cooperators' 'altruism' without themselves contributing to the well-being of the society. In this context A-Life models can provide useful insight both into the general conditions in which cooperation can evolve and into specific forms of cooperation and conflict.
Much work has been done on a specific model of cooperation and conflict known as the 'Iterated Prisoner's Dilemma'. A large number of models have been developed showing the conditions in which it is possible for individuals to cooperate for the common advantage (even when there can be short term gains from non-cooperation)
These abstract models have been applied to a number of concrete situations. Examples include:
The formation of political/military alliances among nations
Price formation in oligopolistic markets
Alliances among competing computer companies during the standards definition process
Bibliography:
Arthur, B., Durlauf S.N., Lane, D.A.,1997, The economy as an evolving complex system, Perseus Books (Investigates the application of A-life methodology and theory to economics)
Axelrod, R. 1997, The Complexity of Cooperation, Princeton University Press (expands on the work presented in the previous reference. Includes a number of empirical examples including the analysis of alliance formation in international relations and the creation and maintenance of alliances among computer companies during the standards setting process)
Axelrod, R.,1984, The evolution of Cooperation, Penguin Books (A classic, completely non-technical book, introducing the 'iterated prisoners dilemma'
Sober E. & Wilson, D.S., 1998, Unto Others, The Evolution and Psychology of Unselfish Behavior, Harvard University Press (an extremely well-written introduction to the biological concept of 'Group Selection' - essential for the definition and understanding of 'altruistic' behavior
Artificial Life as a technology: applications
Goal:
To provide an overview of possible applications of techniques developed in the field of Artificial Life and Evolutionary Computing
Background, techniques and issues:
The techniques and theoretical models currently being used in Artificial Life have given rise to a broad and extremely heterogenous range of applications. These include:
Evolutionary computing in complex optimization problems (e.g. scheduling)
Evolutionary techniques in engineering design (e.g. in aerodynamics)
Artificial evolution of electronic hardware
Evolutionary techniques in the design of telecommunications networks
Combinatorial chemistry (the evolution of new molecules for the pharmaceutical industry)
A-life for the entertainment industry (toys, animation etc.)
A-life modelling as a tool for managing complex systems (traffic, fisheries)
A-Life in education (use of A-life models with primary schoool children and in engineering education)
Bibliography:
Bonabeau E., Theraulaz G., Deneubourg, J-L, 1998, Routing in telecommunications networks with 'Smart' ant-like agents, SFI Working Paper 98-01-003 (an application of 'ant colony' techniques to routing problems in telecommunications)
Calabretta R., Nolfi S. , Parisi D., 1995, An artificial life model for predicting the tertiary structure of unknown proteins that emulates the folding process. In Proceedings of 3.rd European Conference on Artificial Life (ECAL 95), Springer Verlag (shows how A-life techniques can be used to predict the 3D structure of proteins)
D'haeseleer, P., Forrest S., Helman P. , 1996, An immunological approach to change detection. Algorithms, analysis and implications, 1996 IEEE Symposium on Computer Security and Privacy
Dorigo M., Maniezzo V., Colorni A., 1996, The Ant System, Optimization by a colony of cooperating agents. IEEE Transactions on Systems, Man and Cybernetics Part B, 26,1, 29-41.
Grand, S., Cliff D., Malhotra A., 1996, Creatures: artificial life autonomous software agents for home entertainment, Cognitive Science Research Paper, 434, School of Cognitive and Computing Sciences, University of Sussex.
Lund H.H., Paglarini L., Miglino O., (1995), Artistic Design with Genetic Algorithms and Neural Networks, in Alander J.T. (ed.) Proceeedings o 1NWGA, University of Vaasa, Vaasa.
Mcllhagga M., Husbands P., Ives R. (1996), A ison of optimization techniques for integrated manufacturing, ning and scheduling, Cognitive Science Research Paper 408, School of Cognitive and Computing Sciences, University of Sussex
Meczner F., Belew, R.K. , Willuhn W. (1995), Artificial Life applied to adaptive information agents in Knowblock C., Levy A., Chen S-S. & Wiederhold (eds), AAAI Spring Symposium Series: Information gathering from heterogeneous, distributed environments, AAAI
White H., 1988, Economic Prediction using neural networks: the case of IBM daily stock returns, Proceedings of the IEEE International Conference on Neural Networks, San Diego, pp II-451-459 (a classic paper showing how ANN-based prediction techniques may at times be misguided)
Web resources on Artificial Life
Bibliographical sources
Bibliography from the Santa Fe Institute: https://www.santafe.edu/sfi/research/allpubs/index.html
List of web resources produced by the Journal 'Artificial Life': https://alife.santafe.edu/alife/www/index.html
Comprehensive on-line bibliography of A-Life resources from Sussex University: https://www.cogs.susx.ac.uk/users/ezequiel/alife-e/alife.html
On-line bibliography from the Gruppo di Ricerca sulla Vita Artificiale, CNR, Rome: https://kant.irmkant.rm.cnr.it/nolfipub.html
Complexity and Artificial Life Research: https://www.calresco.org/links.htm#alife
Yahoo Artificial Life menu: https://www.yahoo.co.uk/Science/Artificial_Life/
Journals
For a complete bibliography of journals in the field see: https://alife.santafe.edu/alife/journal/index.html
Groups and Institutions
Sussex research group: https://www.cogs.susx.ac.uk/
Santafe Institute: https://www.santafe.edu
Swarm users community e: https://www.santafe.edu/projects/swarm/users/user-community.html
Well-known authors
Beer, Randall: https://www.eecs.cwru.edu/~beer/
Brooks, Rodney: https://www.ai.mit.edu/people/brooks/brooks.html
Burgess, Neil: https://www.icn.ucl.ac.uk/members/Burge12/
Calabretta: https://gracco.irmkant.rm.cnr.it/raffaele/calabrettapub.html
Clark, Andy: https://www.artsci.wustl.edu/~philos/clark_pub.html
Cliff, Dave: https://www.ai.mit.edu/people/davec/davec.html
Dorigo, Marco: https://iridia.ulb.ac.be/~mdorigo/
Kauffman Stuart: https://www.santafe.edu/sfi/People/kauffman/
Langton, Chris: https://www.trail.com/~cgl/
Lund Henrik: https://www.daimi.au.dk/~hhl/
Melanie Mitchell: https://www.santafe.edu/~mm/
Miglino:Orazio: https://gracco.irmkant.rm.cnr.it/orazio/miglino.html
Nolfi, Stefano: https://kant.irmkant.rm.cnr.it/nolfi.html
Parisi, Domenico: https://gracco.irmkant.rm.cnr.it/domenico/dom_bot.htm
Steels, Luc, https://arti9.vub.ac.be/steels/publications.html
|
Privacy
|
© ePerTutti.com : tutti i diritti riservati
:::::
Condizioni Generali - Invia - Contatta
