EvoWorkshops 1993, EvoWorkshops 1994: Progress in Evolutionary Computation pp 61-72 | Cite as
Emergent collective computational abilities in interacting particle systems
Abstract
In recent years computational abilities emerging from systems in nature, especially systems studied in domains such as biology and physics, have attracted much attention from researchers in various fields. In this paper we propose a new computational model based on the interaction of charged particles, and we investigate emergent collective computational abilities of the system. In particle systems, the local motion of particles at the micro level and the state of the system as a whole on the macro level are integrated in a natural way. The local movement of particles is determined by the resultant forces acting on them, and the global system state is described by an energy function. A particle system was constructed to solve Traveling Salesman Problems (TSPs). In comparison with neural networks, this model is able to more effectively make use of the two-dimensional information of city distributions. Finally, to demonstrate the feasibility of our model, we have implemented a simulation of an interacting particle system to solve TSPs on a SUN workstation in C language. The preliminary experimental results show that there are very strong emergent collective computational abilities in interacting particle systems.
Keywords
Position Vector Particle System Travel Salesman Problem Resultant Force Hopfield Neural NetworkPreview
Unable to display preview. Download preview PDF.
References
- 1.Holland, J.: ADAPTATION IN NATURAL AND ARTIFICIAL SYSTEMS. Ann Arbor, The University of Michigan Press (1975)Google Scholar
- 2.Goldberg, D.: GENETIC ALGORITHMS IN SEARCH, OPTIMIZATION, AND MACHINE LEARNING. Addison-Wesley Publishing Company (1989)Google Scholar
- 3.Hopfield, J.: Neural networks and physical systems with emergent collective computational abilities. Proc. Natl. Acad. Sci. USA. 79 (1982) 2554–2558PubMedGoogle Scholar
- 4.Hopfield, J.: Neurons with graded response have collective computational properties like those of two-state neurons. Proc. Natl. Acad. Sci. USA. 81 (1984) 3088–3092PubMedGoogle Scholar
- 5.Hopfield, J.: Neural computation of decisions in optimization problems. Biological Cybernetics. 52 (1985) 141–152PubMedGoogle Scholar
- 6.Fogel, D.: An evolutionary approach to the Traveling Salesman Problem. Biol. Cybernet. 60 (2) (1988) 139–144Google Scholar
- 7.Fogel, D.: Applying evolutionary programming to selected Traveling Salesman Problems. Cybernetics and Systems: An International Journal. 24 (1993) 27–36MathSciNetGoogle Scholar
- 8.Wheeler, J.: Information, physics, quantum: the search for links. Proc. of the 1988 Workshop on Complexity, Entropy, and the Physics of information, Santa Fe, USA. Addison-Wesley Publishing Company (1990) 3–28Google Scholar
- 9.Davies, P.: Why is the physical world so comprehensible. Proc. of the 1988 Workshop on Complexity, Entropy, and the Physics of information, Santa Fe, USA. Addison-Wesley Publishing Company (1990) 61–70Google Scholar
- 10.Toffoli, T.: What are nature's ‘Natural’ ways of computing ? Proc. of the Workshop on Physics and Computation, Dallas, USA. IEEE Computer Society Press (1993) 5–9Google Scholar
- 11.LeÃo, L.: Artificial Physics: the soul of a new discipline. Proc. of the Workshop on Physics and Computation, Dallas, USA. IEEE Computer Society Press (1993) 10–20Google Scholar
- 12.Feynman, R.: Simulating physics with computers. International Journal of Theoretical Physics. 21 (1982) 467–488MathSciNetGoogle Scholar
- 13.Feynman, R.: Quantum mechanical computers. Foundations of Physics. 16 (1986) 507–531Google Scholar
- 14.Feynman, R.: Tiny computers obeying quantum mechanical laws. New Directions in Physics: The Los Alamos 40th Anniversary 7 (1987) 7–25Google Scholar
- 15.Huberman, B.: The ecology of computation. The Ecology of Computation. Elsevier Science Publishers B.V. (1988) 1–4Google Scholar
- 16.Huberman, B., Hogg, T.: The behavior of computational ecologies. The Ecology of Computation. Elsevier Science Publishers B.V. (1988) 77–115Google Scholar
- 17.Farmer, J.: Evolution, Games, and Learning: Models for Adaptation in Machines and Nature. Physica 22D (1986) vii–xiiMathSciNetGoogle Scholar
- 18.Widrow, B.: The rubber mask technique, Parts I and II. Pattern Recognition 5 (1973) 175–211Google Scholar
- 19.Durbin, R., Willshaw, D.: An analogue approach to the travelling salesman problem using an elastic net method. Nature. 326 (1987) 689–691PubMedGoogle Scholar
- 20.Durbin, R., Szeliski, R., Yuille, A.: An analysis of the elastic net approach to the travelling salesman problem. Neural Computation. 1 (1989) 348–358Google Scholar
- 21.Mitchell, M., Crutchfield, J., Hraber, P.: Evolving cellular automata to perform computations: mechanisms and impediments. Physica D (1994) (to appear)Google Scholar
- 22.Das, R., Mitchell, M., Crutchfield, J.: A genetic algorithm discovers particle-based computation in cellular automata. Third Parallel Problem-Solving From Nature Conference. (1994) (to appear)Google Scholar
- 23.Crutchfield, J., Mitchell, M.: The evolution of emergent computation. Science. (1994) (to appear)Google Scholar