Abstract
This paper addresses the problem of tuning hyperparameters in support vector machine modeling. A Genetic Algorithm-based wrapper, which seeks to evolve hyperparameter values using an empirical error estimate as a fitness function, is proposed and experimentally evaluated on a medical dataset. Model selection is then fully automated. Unlike other hyperparameters tuning techniques, genetic algorithms do not require supplementary information making them well suited for practical purposes. This approach was motivated by an application where the number of parameters to adjust is greater than one. This method produces satisfactory results.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Ayat, N., Cheriet, M., Suen, C.: Empirical error based optimization of SVM kernels (2002)
Burges, C.J.C.: A tutorial on support vector machines for pattern recognition. Data Mining and Knowledge Discovery, 121–167 (1998)
Chapelle, O., Vapnik, V., Bousquet, O., Mukherjee, S.: Choosing multiple parameters for support vector machines. Machine Learning 46(1), 131–159 (2002)
Cohen, G., Hilario, M., Sax, H., Hugonnet, S.: Asymmetrical margin approach to surveillance of nosocomial infections using support vector classification. In Intelligent Data Analysis in Medicine and Pharmacology (2003)
Cortes, C., Vapnik, V.: Support vector networks. Machine Learning 20(3), 273–297 (1995)
Cristianini, N., Taylor, J.S.: An Introduction to Support Vector Machines. Cambridge University Press, Cambridge (2000)
Duan, K., Keerthi, S.S., Poo, A.N.: Evaluation of simple performance measures for tuning svm hyperparameters. In Technical Report CD-01-11, Department of Mechanical Engineering. National University of Singapore, Singapore (2001)
Fletcher, R.: Practical Methods of Optimization. John Wiley and Sons, Chichester (1987)
Goldberg, D.E.: Genetic Algorithms in Search, Optimization, and Machine Learning. Addison-Wesley, Reading (1989)
Harbarth, S., Ruef, C., Francioli, P., Widmer, A., Pittet, D.: Swiss-Noso Network. Nosocomial infections in Swiss university hospitals: a multi-centre survey and review of the published experience. Schweiz Med. Wochenschr 129, 1521–1528 (1999)
Joachims, T.: Estimating the generalization performance of a SVM efficiently. In: Langley, P. (ed.) Proceedings of ICML 2000, 17th International Conference on Machine Learning, Stanford, US, pp. 431–438. Morgan Kaufmann Publishers, San Francisco (2000)
Karakoulas, G., Shawe-Taylor, J.: Optimizing classifiers for imbalanced training sets. In: Advances in Neural Information Processing Systems (NIPS 1999), MIT Press, Cambridge (1999)
Keerthi, S.S.: Efficient tuning of svm hyperparameters using radius/margin bound and iterative algorithms. In: IEE Transactions on Neural Networks, pp. 1225–1229 (2003)
Luntz, A., Brailowsky, V.: On estimation characters obtained in statistical procedure of recognition. Technicheskaya Kibernetica 3 (1969) (in russian)
Vapnik, V.: Statistical Learning Theory. Wiley, Chichester (1998)
Veropoulos, K., Cristianini, N., Campbell, C.: Controlling the sensitivity of support vector machines. In: Proc. International Joint Conference on Artificial Intelligence (1999)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2004 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Cohen, G., Hilario, M., Geissbuhler, A. (2004). Model Selection for Support Vector Classifiers via Genetic Algorithms. An Application to Medical Decision Support. In: Barreiro, J.M., MartÃn-Sánchez, F., Maojo, V., Sanz, F. (eds) Biological and Medical Data Analysis. ISBMDA 2004. Lecture Notes in Computer Science, vol 3337. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-30547-7_21
Download citation
DOI: https://doi.org/10.1007/978-3-540-30547-7_21
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-23964-2
Online ISBN: 978-3-540-30547-7
eBook Packages: Springer Book Archive