Abstract
We consider the problem of multi-modal regression estimation under the assumption that a kernel-based approach is applicable within each particular modality. The Cartesian product of the linear spaces into which the respective kernels embed the output scales of single sensors is employed as an appropriate joint scale corresponding to the idea of combining modalities at the sensor level. This contrasts with the commonly adopted method of combining classifiers inferred from each specific modality. However, a significant risk in combining linear spaces is that of overfitting. To address this, we set out a stochastic method for encompassing modal-selectivity that is intrinsic to (that is to say, theoretically contiguous with) the selected kernel-based approach.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
It is incorrect to speak about strictly normal densities since the dimensionality of each linear space \( \tilde{\Omega }_{i} \) depends on the respective kernel function. As a result, the normalization coefficient of any density \( \left( {\uppsi_{i} (a_{i} ),\;a_{i} \, \in \,\tilde{\Omega }_{i} } \right) \) cannot be specified before the kernel is completely defined.
References
Vapnik, V.: Estimation of Dependencies Based on Empirical Data. Springer, New York (1982)
Schölkopf, B., Smola, A.J.: Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond. MIT Press, Cambridge (2001)
Bach, F.R., Lanckriet, G.R., Jordan, M.I.: Multiple kernel learning, conic duality, and the SMO algorithm. In: Proceedings of the Twenty-First International Conference on Machine Learning, p. 6. ACM (2004)
Fan, J., Lv, J.: A selective overview of variable selection in high dimensional feature space. Statistica Sinica 20, 101–148 (2010)
Tatarchuk, A., Sulimova, V., Windridge, D., Mottl, V., Lange, M.: Supervised selective combining pattern recognition modalities and its application to signature verification by fusing on-line and off-line kernels. In: Benediktsson, J.A., Kittler, J., Roli, F. (eds.) MCS 2009. LNCS, vol. 5519, pp. 324–334. Springer, Heidelberg (2009)
Vapnik, V.: Statistical Learning Theory. Wiley, New York (1998)
Mottl, V.: Metric spaces admitting linear operations and inner product. Dokl. Math. 67(1), 140–143 (2003)
Mottl, V., Krasotkina, O., Seredin, O., Muchnik, I.: Kernel fusion and feature selection in machine learning. In: Proceedings of the 8th IASTED International Conference on Intelligent Systems and Control, Cambridge, USA (2005)
Mottl, V., Krasotkina, O., Seredin, O., Muchnik, I.B.: Principles of multi-kernel data mining. In: Perner, P., Imiya, A. (eds.) MLDM 2005. LNCS (LNAI), vol. 3587, pp. 52–61. Springer, Heidelberg (2005)
Mottl, V.V., Seredin, O.S., Krasotkina, O.V., Muchnik, I.B.: Fusing of potential functions in reconstructing dependences from empirical data. Dokl. Math. 71(2), 315–319 (2005). MAIK Nauka/Interperiodica
Mottl, V., Tatarchuk, A., Sulimova, V., Krasotkina, O., Seredin, O.: Combining pattern recognition modalities at the sensor level via kernel fusion. In: Haindl, M., Kittler, J., Roli, F. (eds.) MCS 2007. LNCS, vol. 4472, pp. 1–12. Springer, Heidelberg (2007)
Acknowledgements
This research is funded by RFBR, grants 14-07-00964, 14-07-00527.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2015 Springer International Publishing Switzerland
About this paper
Cite this paper
Krasotkina, O., Seredin, O., Mottl, V. (2015). Supervised Selective Combination of Diverse Object-Representation Modalities for Regression Estimation. In: Schwenker, F., Roli, F., Kittler, J. (eds) Multiple Classifier Systems. MCS 2015. Lecture Notes in Computer Science(), vol 9132. Springer, Cham. https://doi.org/10.1007/978-3-319-20248-8_8
Download citation
DOI: https://doi.org/10.1007/978-3-319-20248-8_8
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-20247-1
Online ISBN: 978-3-319-20248-8
eBook Packages: Computer ScienceComputer Science (R0)