Abstract
This position paper proposes a mathematical modeling approach for a certain class of connectionist network structures. Investigation of the structure of an artificial neural network (ANN) in that class (paradigm) suggested the use of geometric and categorical modeling methods in the following sense. A (noncommutative) geometric space can be interpreted as a so-called geometric net. To a given ANN a corresponding geometric net can be associated. Geometric spaces form a category. Consequently, one obtains a category of geometric nets with a suitable notion of morphism. It is natural to interpret a learning step of an ANN as a morphism, thus learning corresponds to a finite sequence of morphisms (the associated networks are the objects). An associated (“local”) geometric net is less complex than the original ANN, but it contains all necessary information about the network structure. The association process together with learning (expressed by morphisms) leads to a commutative diagram corresponding to a suitable natural transformation. Commutativity can be exploited to make learning “cheaper”. The simplified mathematical network model was used in ANN simulation applied in an industrial project on quality control. The “economy” of the model could be observed in a considerable increase of performance and decrease of production costs.
This is a preview of subscription content, log in via an institution to check access.
Preview
Unable to display preview. Download preview PDF.
Similar content being viewed by others
References
[AHS90] J. Adámek, H. Herrlich, and G.E. Strecker. Abstract and Concrete Categories. John Wiley & Sons, 1990.
[And88] J. André. Endliche nichtkommutative Geometrie. Annales Univ. Saraviensis. Ser. Math., 2(1):1–136, 1988.
[And92] J. André. Configurational conditions and digraphs. Journal of Geometry, 43:22–29, 1992.
[And93] J. André. On non-commutative geometry. Annales Univ. Saraviensis. Ser. Math., 4(2):93–129, 1993.
[Eck90] R. Eckmiller. Concerning the emerging role of geometry in neuroinformatics. In Parallel Processing in Neural Systems and Computers. R.Eckmiller, G.Hartmann, G.Hauske (Eds.), North-Holland, 1990.
[Gei94] H. Geiger. Optical quality control with selflearning systems using a combination of algorithmic and neural network approaches. Proceedings of the Second European Congress on Intelligent Techniques and Soft Computing, EUFIT'94, Aachen, September 20-23, 1994.
[GP] H. Geiger and J. Pfalzgraf. Modeling a connectionist network paradigm: Geometric and categorical perspectives. In preparation.
[GP95] H. Geiger and J. Pfalzgraf. Quality control connectionist networks supported by a mathematical model. In Proceedings of the International Conference on Engineering Applications of Artificial Neural Networks (EANN'95), 21-23 August 1995, Helsinki. A.B.Bulsari, S.Kallio(Editors), Finnish AI Society, 1995.[Lan98] S. Mac Lane. Categories for the Working Mathematician. Springer Verlag, Graduate Texts in Mathematics 5, 2nd ed., 1998.
[LS96] F.W.Lawvere and S.H. Schanuel. Conceptual Mathematics: A First Introduction to Categories. Cambridge University Press, 1996.
[NKK94] D. Nauck, F. Klawonn, and R. Kruse. Neuronale Netze und Fuzzy-Systeme. Vieweg Verlag, 1994.
[Pfa85] J. Pfalzgraf. On a model for noncommutative geometric spaces. Journal of Geometry, 25:147–163, 1985.
[Pfa87] J. Pfalzgraf. A note on simplices as geometric configurations. Archiv der Mathematik, 49:134–140, 1987.
[Pfa94] J. Pfalzgraf. On a general notion of a hull. In Automated Practical Reasoning, J. Pfalzgraf and D. Wang(nteds.). Texts and Monographs in Symbolic Computation, Springer Verlag Wien, New York, 1994.
[Pfa95] J. Pfalzgraf. Graph products of groups and group spaces. J. of Geometry, 53:131–147, 1995.
[Pfa98] J. Pfalzgraf. On a category of geometric spaces and geometries induced by group actions. Ukrainian Jour. Physics, 43,7:847–856, 1998.
[Pie91] B.C. Pierce. Basic Category Theory for Computer Scientists. The MIT Press, Cambridge, Massachusetts, 1991.
[Roj93] R. Rojas. Theorie der neuronalen Netze. Springer Verlag, 1993. Springer-Lehrbuch.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2001 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Pfalzgraf, J. (2001). A Note on Modeling Connectionist Network Structures: Geometric and Categorical Aspects. In: Campbell, J.A., Roanes-Lozano, E. (eds) Artificial Intelligence and Symbolic Computation. AISC 2000. Lecture Notes in Computer Science(), vol 1930. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-44990-6_14
Download citation
DOI: https://doi.org/10.1007/3-540-44990-6_14
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-42071-2
Online ISBN: 978-3-540-44990-4
eBook Packages: Springer Book Archive