Abstract
The approximation capability of support vector machines (SVMs) is investigated. We show the universal approximation capability of SVMs with various kernels, including Gaussian, several dot product, or polynomial kernels, based on the universal approximation capability of their standard feedforward neural network counterparts. Moreover, it is shown that an SVM with polynomial kernel of degree p − 1 which is trained on a training set of size p can approximate the p training points up to any accuracy.
Similar content being viewed by others
References
Barron, A. R.: Universal approximation bounds for superpositions of a sigmoidal function, IEEE Transactions on Information Theory 39 (1993), 930–945.
Ben-David, S., Eiron, N. and Simon, H. U.: Limitations of Learning via Embedding in Euclidean Half Spaces. NeuroCOLT Technical Report 2001-086, 2001.
Burges, C. J. C.: A tutorial on support vector machines for pattern recognition, Data Mining and Knowledge Discovery 2(2) (1998), 121–167.
Cortes, C. and Vapnik, V.: Support vector network, Machine Learning 20 (1995), 1–20.
Cybenko, G.: Approximation by superposition of a sigmoidal function, Mathematics of Control, Signals, and Systems 2 (1989), 303–314.
Funahashi, K.: On the approximate realization of continuous mappings by neural networks, NeuralNetworks 2 (1989), 183–192.
Gurvits, L.: A note on a scale-sensitive dimension of linear bounded functionals in Banach spaces, In: M. Li and A. Maruoska (eds.), Algorithmic Learning Theory ALT-97, Springer, 1997, pp. 352–363.
Hecht-Nielsen, R.: Kolmogorov's mapping neural network existence theorem, In: InternationalConference on NeuralNetworks, vol. 3, IEEE, Washington DC, 1989, pp.11–14.
Hornik, K.: Some new results on neural network approximation, NeuralNetworks 6 (1993), 1069–1072.
Hornik, K., Stinchcombe, M. and White, H.: Multilayer feedforward networks are universal approximators, NeuralNetworks 2 (1989), 359–366.
Joachims, T.: Text categorization with support vector machines: Learning with many relevant features, In: Proceedings of the European Conference on Machine Learning, Springer, Berlin, 1998, pp. 137–142.
Kurkova, V., Kainen, P. C. and Kreinovich, V.: Estimates of the number of hidden neurons and variation with respect to half-spaces, NeuralNetworks 10 (1997), 1061–1068.
Müller, K.-R., Smola, A., Rätsch, G., Schölkopf, B., Kohlmorgen, J. and Vapnik, V.: Predicting time series with support vector machines, In: W. Gerstner, A. Germond, M. Hasler and J.-D. Nicoud (eds.), ArtificialNeuralNetworks — ICANN '97, Springer, 1997, pp. 999–1004.
Park, J. and Sandberg, I. W.: Universal approximation using radial-basis-function networks, NeuralComputation 3(2) (1991), 246–257.
Park, J. and Sandberg, I. W.: Approximation and radial-basis-function networks, Neural Computation 5 (1993), 305–316.
Pinkus, A.: Approximation theory of the MLP model in neural networks, Acta Numerica, (1999), 143–195.
Scarselli, F. and Tsoi, A. C.: Universal approximation using feedforward neural networks: A survey of some existing methods, and some new results, Neural Networks 11(1) (1998), 15–37.
Shawe-Taylor, J., Bartlett, P. L., Williamson, R. and Anthony, M.: Structural risk minimization over data dependent hierarchies. Technical report, NeuroCOLT, 1996.
Schölkopf, B. and Smola, A. J.: Learning with kernels, The MIT Press, 2001.
Smola, A.: Support Vector Machines, Tutorial at ICANN '01, Vienna, Austria, 2001.
Vapnik, V. and Chervonenkis, A.: Theory of Pattern Recognition, Nauka, Moscow, 1974.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Hammer, B., Gersmann, K. A Note on the Universal Approximation Capability of Support Vector Machines. Neural Processing Letters 17, 43–53 (2003). https://doi.org/10.1023/A:1022936519097
Issue Date:
DOI: https://doi.org/10.1023/A:1022936519097