Abstract
We propose cutting plane algorithm for solving the linear pattern separation problem, which is a particular case of the more general topic of data mining. The solution we provided, based on convex programming, can also be applied to any other pattern separation scheme based on a convex discriminant like linear piecewise or quadratic models. Some experimentations are reported with different large databases together with a comparison with a direct implemention with two commercial specialized codes.
Chapter PDF
Similar content being viewed by others
References
Bosh, R.A., Smith, J.A.: Separating Hyperplanes and the Authorship of the Disputed Federalist Papers. American Mathematical Monthly 105(7), 601–608 (1995)
Ferris, M.C., Munson, T.S.: Interior Point Methods for Massive Support Vector Machines, Cours/Sèminaire du 3e cycle romand de recherche opèrationnelle, Zinal, Switzerland (March 2001)
Han, J., Kamber, M.: Data Mining: Concept and Techniques. Morgan Kaufmann Publishers, San Francisco (2000)
Mangasarian, O.L.: Linear and Non-linear Separation of Patterns by linear programming. Operations Research 13, 444–452
Mangasarian, O.L., Setino, R., Wolberg, W.: Pattern Recognition via linear programming: Theory and Applications to Medical Diagnosis (1990)
Viveros, M.S., Nearhos, J.P., Rothman, M.J.: Applying Data Mining Techniques to a Health Insurance Information System. In: 22nd VLDB Conference, Mumbai (Bombay), India, pp. 286–294 (1996)
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
Tadonki, C., Vial, JP. (2004). Efficient Algorithm for Linear Pattern Separation. In: Bubak, M., van Albada, G.D., Sloot, P.M.A., Dongarra, J. (eds) Computational Science - ICCS 2004. ICCS 2004. Lecture Notes in Computer Science, vol 3036. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-24685-5_118
Download citation
DOI: https://doi.org/10.1007/978-3-540-24685-5_118
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-22114-2
Online ISBN: 978-3-540-24685-5
eBook Packages: Springer Book Archive