Abstract
Multilayer Perceptrons (MLPs) use scalar products to compute weighted activation of neurons providing decision borders using combinations of soft hyperplanes. The weighted fun-in activation function may be replaced by a distance function between the inputs and the weights, offering a natural generalization of the standard MLP model. Non-Euclidean distance functions may also be introduced by normalization of the input vectors into an extended feature space. Both approaches influence the shapes of decision borders dramatically. An illustrative example showing these changes is provided.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Bishop, C.: Neural Networks for Pattern Recognition. Clarendon Press, Oxford, 1995.
Schiffman, W., Joost, M. and Werner, R.: Comparison of optimized backpropagation algorithms, Proc. of ESANN'93, Brussels 1993, pp. 97–104.
Duch, W. and Jankowski, N.: New neural transfer functions, Applied Mathematics and Computer Science 7 (1997), 639–658.
Duch, W.: Neural minimal distance methods, Proc. 3rd Conf. on Neural Networks and Their Applications, Kule, Poland, Oct. 14–18, 1997, pp. 183–188.
Duch, W., Grudzinski, K. and Diercksen, G. H. F.: Minimal distance neural methods. World Congress of Computational Intelligence, May 1998, Anchorage, Alaska, IJCNN'98 Proceedings, pp. 1299–1304.
Duch, W., Adamczak, R. and Diercksen, G. H. F.: Distance-based multilayer perceptrons, In: M. Mohammadian (ed.), Computational Intelligence for Modelling Control and Automation. Neural Networks and Advanced Control Strategies, IOS Press, Amsterdam, pp. 75–80.
Krishnaiah, P. R. and Kanal, L. N. (eds): Handbook of statistics 2: classification, pattern recognition and reduction of dimensionality (North Holland, Amsterdam 1982).
Wasserman, P. D.: Advanced Methods in Neural Networks, van Nostrand Reinhold, 1993.
Kohonen, T.: Self-Organizing Maps, Berlin, Springer-Verlag, 1995.
Reilly, D. L., Cooper, L. N. and Elbaum, C.: A neural model for category learning, Biological Cybernetics 45 (1982), 35–41.
Duch, W. and Diercksen, G. H. F.: Feature Space Mapping as a universal adaptive system, Comp. Phys. Communic. 87 (1995), 341–371.
Lippmann, R. P.: An introduction to computing with neural nets, IEEE Magazine on Acoustics, Signal and Speech Processing 4 (1987), 4–22.
Floreen, P.: The convergence of Hamming memory networks, Trans. Neural Networks 2 (1991), 449–457.
Wilson, D. R. and Martinez, T. R.: Improved heterogenous distance functions. J. Artificial Intelligence Research 6 (1997), 1–34.
Duch, W., Adamczak, R. and Jankowski, N.: Initialization and optimization of multilayered perceptrons, 3rd Conf. on Neural Networks and Their Applications, Kule, Poland, October 1997, pp. 105–110.
Duch, W., Adamczak, R., Grabczewski, K. and Żal, G.: Hybrid neural-global minimization method of logical rule extraction, Journal of Advanced Computational Intelligence (in print).
Ridella, S., Rovetta, S. and Zunino, R.: Circular backpropagation networks for classification, IEEE Trans. Neural Networks 8 (1997), 84–97.
Kirby, M. J. and Miranda, R.: Circular nodes in neural networks, Neural Computations 8 (1996), 390–402.
Dorffner, G.: A unified framework for ofMLPs and RBFNs: introducing conic section function networks, Cybernetics & Systems 25 (1994), 511–554.
Amari, S.-I.: Information geometry of the EM and em algorithms for neural networks, Neural Networks 8 (1995), 1379–1408.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Duch, W., Adamczak, R. & Diercksen, G.H. Neural Networks in Non-Euclidean Spaces. Neural Processing Letters 10, 201–210 (1999). https://doi.org/10.1023/A:1018728407584
Issue Date:
DOI: https://doi.org/10.1023/A:1018728407584