Abstract
The architecture of a neural network can be optimized by structure evolution. The structure evolution is based upon a two-stage evolution strategy (multipopulation strategy): On the population level, the structure is optimized, on the individual level, the parameters are adapted. For the variation of the (discrete) architecture, a mutation operator must be defined. To attain successful optimization, a mutation operator must satisfy two main conditions in the space of structures: First the principle of strong causality must be obeyed (smoothness of the fitness landscape in the space of structure), and second, a transition path between the structures must be guaranteed. In this paper different heuristic mutation operators will be defined and examined on their behavior with respect to strong causality and to neighborhood relation in the space of structures.
Preview
Unable to display preview. Download preview PDF.
References
Angeline, P. J., Saunders, G. M., Pollack, J. B.: An Evolutionary Algorithm that Constructs Recurrent Neural Networks; IEEE Transactions on Neural Networks, vol. 5, no. 1, pp. 54–65; 1994
Born, J., Santibáñez-Koref, I., Voigt, H.-M.: Designing Neural networks by Adaptively Building Blocks in Cascades; Proc. of the 3. Conf. on Parallel Problem Solving from Nature; Jerusalem; 1994
Fogel, D. B.: Using Evolutionary Programming to Create Neural Networks that are Capable of Playing Tic-Tac-Toe; IEEE Int. Conf. on Neural Networks, vol. 2, pp. 875–880; San Francisco, CA; 1993
Fogel, D. B.: An Introduction to Simulated Evolutionary Optimization; IEEE Transactions on Neural Networks, vol. 5, no. 1, pp. 3–14; 1994
Fogel, L. J., Owens, A. J., Walsh, M. J.: Artificial Intelligence Through Simulated Evolution; John Wiley & Sons; 1966
Goldberg, D. E.: Genetic Algorithms in Search Optimization and Machine Learning; Addison Wesley; Reading, MA; 1989
Gruau, F.: Genetic Synthesis of Modular Neural Networks; Proc. of the 5. Int. Conf. on Genetic Algorithms; Morgan Kaufmann Publishers; San Mateo, CA; 1993
Gruau, F., Whitley, D.: Adding Learning to the Cellular Development of Neural Networks: Evolution and the Baldwin Effect; Evolutionary Computation, vol. 1, no. 3; The Massachusetts Institute of Technology; 1993
Holland, J. H.: Adaptation in Natural and Artificial Systems; The University of Michigan Press; Ann Arbor, MI; 1975
Jacob, C., Rehder, J.: Evolution of Neural Net Architectures by a hierarchical grammar-based Genetic System; Artificial Neural Nets and Genetic Algorithms; Proc. of the Int. Conf. in Innsbruck; Springer-Verlag; Wien; 1993
Koza, J. R., Rice, J. P.: Genetic Generation of both the Weights and Architecture for a Neural Network; Proc. of the Int. Joint Conf. on Neural Networks, Seattle; IEEE Press vol. 2, pp. 47–56; 1991
Kitano, H.: Designing neural networks using genetic algorithm with graph generating system; Complex Systems, vol. 4, pp. 461–476; 1992
Lohmann, R.: Selforganisation by Evolution Strategy in Visual Systems; in: Voigt, Mühlenbein, Schwefel (ed.): Evolution and Optimization '89; pp. 61–68; Akademie Verlag; Berlin; 1990
Lohmann, R.: Bionische Verfahren zur Entwicklung visueller Systeme; Master thesis, Technical University Berlin, Bionics and Evolution Techniques; 1991
Lohmann, R.: Structure Evolution and Incomplete Induction; Proc. of the 2. Conf. on Parallel Problem Solving from Nature, pp. 175–185; Brussels; 1992
Maniezzo, V.: Genetic Evolution of the Topology and Weight Distribution of Neural Networks; IEEE Transactions on Neural Networks, vol. 5, no. 1, pp. 39–53; 1994
McDonnell, J. R., Waagen, D.: Evolving Neural Network Connectivity; IEEE Int. Conf. on Neural Networks, vol. 2, pp. 863–868; San Francisco, California; 1993
McDonnell, J. R., Waagen, D.: Evolving Recurrent Perceptrons for Time-Series Modeling; IEEE Transactions on Neural Networks, vol. 5, no. 1, pp. 24–38; 1994
Miller, G., Todd, P., Hedge, S.: Designing Neural networks using genetic Algorithms; Proc. of the 3. Int. Conf. on Genetic Algorithms, pp. 379–384; 1989
Rechenberg, I.: Evolutionsstrategie — Optimierung technischer Systeme nach Prinzipien der biologischen Evolution; Frommann-Holzboog; Stuttgart; 1973
Rechenberg, I.: Evolutionsstrategie — Optimierung nach Prinzipien der biologischen Evolution; in: Albertz, J. (ed.): Evolution und Evolutionsstrategien in Biologie, Technik und Gesellschaft; Freie Akademie; Wiesbaden; 1989
Rechenberg, I., Gawelczyk, A., Görne, T., Hansen, N., Herdy, M., Kost, B., Lohmann, R., Ostermeier, A., Trint, K., Utecht, U.: Evolutionsstrategische Strukturbildung in neuronalen Systemen; Statusseminar des BMFT Neuroinformatik, Maurach; ed: Projektträger Informationstechnik BMFT, DLR, pp. 25–34; 1992
Rechenberg, I.: Evolutionsstrategie '94; Frommann-Holzboog; Stuttgart; 1994
Rumelhart, D. E., McClelland, J. L.: Parallel Distributed Processing, Vol. 1 & 2; MIT-Press; Cambridge, MA; 1986
Schwefel, H.-P.: Numerische Optimierung von Computer-Modellen mittels der Evolutionsstrategie; Birkhäuser; Basel, Stuttgart; 1977
Trint, K., Utecht, U.: Methodik der Strukturevolution; Technical Report; Technical University Berlin, Bionics and Evolution Techniques; 1992
Trint, K., Utecht, U.: Selection Mechanisms in Two-Stage Evolution Strategies for Noisy Evaluations; Technical Report TR-94-04; Technical University Berlin, Bionics and Evolution Techniques; 1994
Whitley, D., Starkweather, T., Bogart, C.: Genetic algorithms and neural networks: optimizing connections and connectivity; Parallel Computing, vol. 14, pp. 347–361; 1990
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1994 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Utecht, U., Trint, K. (1994). Mutation operators for structure evolution of neural networks. In: Davidor, Y., Schwefel, HP., Männer, R. (eds) Parallel Problem Solving from Nature — PPSN III. PPSN 1994. Lecture Notes in Computer Science, vol 866. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-58484-6_292
Download citation
DOI: https://doi.org/10.1007/3-540-58484-6_292
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-58484-1
Online ISBN: 978-3-540-49001-2
eBook Packages: Springer Book Archive