Abstract
In two-class problems, the linear combination of the outputs (scores) of an ensemble of classifiers is widely used to attain high performance. In this paper we investigate some techniques aimed at dynamically estimate the coefficients of the linear combination on a pattern per pattern basis. We will show that such a technique allows providing better performance than those of static combination techniques, whose parameters are computed beforehand. The coefficients of the linear combination are dynamically computed according to the Wilcoxon-Mann-Whitney statistic. Reported results on a multi-modal biometric dataset show that the proposed dynamic mechanism allows attaining very low error rates when high level of precision are required.
Chapter PDF
Similar content being viewed by others
References
Fawcett, T.: An introduction to ROC Analysis. Pattern Recognition Letters 27(8), 861–874 (2006)
Fiérrez-Aguilar, J., Chen, Y., Ortega-Garcia, J., Jain, A.K.: Incorporating Image Quality in Multi-algorithm Fingerprint Verification. In: Zhang, D., Jain, A.K. (eds.) ICB 2005. LNCS, vol. 3832, pp. 213–220. Springer, Heidelberg (2005)
Fumera, G., Roli, F.: A theoretical and experimental analysis of linear combiners for multiple classifier systems. IEEE Transactions on Pattern Analysis and Machine Intelligence 27, 942–956 (2005)
Hanley, J.A., McNeil, B.J.: The meaning and the use of the area under a receiver operanting charateristic curve. Radiology 143, 29–36 (1982)
Huang, J., Ling, C.: Using AUC and Accuracy in Evaluating Learning Algorithms. IEEE Transactions on Knowledge and Data Engineering 17, 299–310 (2005)
Kittler, J., Hatef, M., Duin, R., Matas, J.: On combing classifiers. IEEE Trans. on PAMI 20(3), 226–239 (1998)
Kuncheva, L.: Combining Pattern Classifiers: Methods and Algorithms. John Wiley & Sons Inc., Chichester (2004)
Marcialis, G.L., Roli, F.: Fusion of multiple fingerprint matchers by single layer perceptron with class-separation loss function. Pattern Recognition Letters 26, 1830–1839 (2005)
Nandakumar, K., Jain, A., Dass, S.: Quality-based Score Level Fusion in Multibiometric Systems. In: ICPR 2006, pp. 473–476 (2006)
Poh, N., Bengio, S.: Improving Fusion with Margin-Derived Confidence In Biometric Authentication Tasks. In: Kanade, T., Jain, A., Ratha, N.K. (eds.) AVBPA 2005. LNCS, vol. 3546, pp. 474–483. Springer, Heidelberg (2005)
Poh, N., Bengio, S.: Database, protocols and tools for evaluating score-level fusion algorithms in biometric authentication. Pattern Recognition 39(2), 223–233 (2006)
Ross, A.A., Nandakumar, K., Jain, A.K.: Handbook of multibiometrics. Springer, Heidelberg (2006)
Tronci, R., Giacinto, G., Roli, F.: Dynamic Score Combination: A Supervised and Unsupervised Score Combination. In: Perner, P. (ed.) MLDM 2009. LNCS (LNAI), vol. 5632, pp. 163–177. Springer, Heidelberg (2009)
Tumer, K., Gosh, J.: Linear and order statistics combiners for pattern classification. In: Combining Artificial Neural Nets, pp. 127–162. Springer, Heidelberg (1999)
Marrocco, C., Duin, R.P.W., Tortorella, F.: Maximizing the area under the ROC curve by pairwise feature combination. Pattern Recognition 41(6), 1961–1974 (2008)
Bigun, E., Bigun, J., Duc, B., Fischer, S.: Expert conciliation for multi modal person authentication systems by Bayesian statistics. In: Bigün, J., Borgefors, G., Chollet, G. (eds.) AVBPA 1997. LNCS, vol. 1206, pp. 291–300. Springer, Heidelberg (1997)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2010 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Lobrano, C., Tronci, R., Giacinto, G., Roli, F. (2010). Dynamic Linear Combination of Two-Class Classifiers. In: Hancock, E.R., Wilson, R.C., Windeatt, T., Ulusoy, I., Escolano, F. (eds) Structural, Syntactic, and Statistical Pattern Recognition. SSPR /SPR 2010. Lecture Notes in Computer Science, vol 6218. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-14980-1_46
Download citation
DOI: https://doi.org/10.1007/978-3-642-14980-1_46
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-14979-5
Online ISBN: 978-3-642-14980-1
eBook Packages: Computer ScienceComputer Science (R0)