Abstract
Kernelized learning algorithms have seen a steady growth in popularity during the last decades. The procedure to estimate the performances of these kernels in real applications is typical computationally demanding due to the process of hyper-parameter selection. This is especially true for graph kernels, which are computationally quite expensive. In this paper, we study an approach that substitutes the commonly adopted procedure for kernel hyper-parameter selection by a multiple kernel learning procedure that learns a linear combination of kernel matrices obtained by the same kernel with different values for the hyper-parameters. Empirical results on real-world graph datasets show that the proposed methodology is faster than the baseline method when the number of parameter configurations is large, while always maintaining comparable and in some cases superior performances.
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
Notes
- 1.
A graph where vertices are atoms and edges are chemical bonds; the label attached to each vertex reports the atom type.
- 2.
References
Aiolli, F., Da San Martino, G., Sperduti, A.: A kernel method for the optimization of the margin distribution. In: KĹŻrková, V., Neruda, R., KoutnĂk, J. (eds.) ICANN 2008, Part I. LNCS, vol. 5163, pp. 305–314. Springer, Heidelberg (2008)
Aiolli, F., Donini, M.: EasyMKL: a scalable multiple kernel learning algorithm. Neurocomputing 169, 215–224 (2015)
Aiolli, F., Donini, M., Navarin, N., Sperduti, A.: Multiple graph-kernel learning. In: IEEE SSCI, Cape Town, pp. 1607–1614. IEEE (2015)
Da San Martino, G., Navarin, N., Sperduti, A.: A tree-based kernel for graphs. In: SDM, pp. 975–986 (2012)
Da San Martino, G., Navarin, N., Sperduti, A.: Ordered decompositional, DAG kernel enhancements. Neurocomputing 192, 92–103 (2016)
Da San Martino, G., Sperduti, A.: Mining structured data. IEEE Comput. Intell. Mag. 5(1), 42–49 (2010)
Dobson, P.D., Doig, A.J.: Distinguishing enzyme structures from non-enzymes without alignments. J. Mol. Biol. 330(4), 771–783 (2003)
Gnen, M., Alpaydin, E.: Multiple kernel learning algorithms. JMLR 12, 2211–2268 (2011)
Hastie, T., Rosset, S., Tibshirani, R., Zhu, J.: The entire regularization path for the support vector machine. JMLR 5(2), 1391–1415 (2004)
Helma, C., Cramer, T., Kramer, S., Raedt, L.D.: Data mining and machine learning techniques for the identification of mutagenicity inducing substructures and structure activity relationships of noncongeneric compounds. J. Chem. Inf. Model. 44(4), 1402–1411 (2004)
Navarin, N., Sperduti, A., Tesselli, R.: Extending local features with contextual information in graph kernels. In: Arik, S., Huang, T., Lai, W.K., Liu, Q. (eds.) ICONIP 2015. LNCS, vol. 9492, pp. 271–279. Springer, Heidelberg (2015). doi:10.1007/978-3-319-26561-2_33
Shawe-Taylor, J., Cristianini, N.: Kernel Methods for Pattern Analysis. Cambridge University Press, New York (2004)
Shervashidze, N., Borgwardt, K.M.: Fast subtree kernels on graphs. In: Bengio, Y., Schuurmans, D., Lafferty, J.D., Williams, C.K.I., Culotta, A. (eds.) NIPS, pp. 1660–1668. Curran Associates Inc., Red Hook (2009)
Tesselli, R.: Adding contextual information to graph kernels. Master’s thesis, Università di Padova (2015)
Wale, N., Watson, I.A., Karypis, G.: Comparison of descriptor spaces for chemical compound retrieval and classification. Knowl. Inf. Syst. 14(3), 347–375 (2008)
Weislow, O.S., Kiser, R., Fine, D.L., Bader, J., Shoemaker, R.H., Boyd, M.R.: New soluble-formazan assay for HIV-1 cytopathic effects: application to high-flux screening of synthetic and natural products for AIDS-antiviral activity. J. Natl. Cancer Inst. 81(8), 577–586 (1989)
Acknowledgments
This work was supported by the University of Padova under the strategic project BIOINFOGEN.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer International Publishing AG
About this paper
Cite this paper
Massimo, C.M., Navarin, N., Sperduti, A. (2016). Hyper-Parameter Tuning for Graph Kernels via Multiple Kernel Learning. In: Hirose, A., Ozawa, S., Doya, K., Ikeda, K., Lee, M., Liu, D. (eds) Neural Information Processing. ICONIP 2016. Lecture Notes in Computer Science(), vol 9948. Springer, Cham. https://doi.org/10.1007/978-3-319-46672-9_25
Download citation
DOI: https://doi.org/10.1007/978-3-319-46672-9_25
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-46671-2
Online ISBN: 978-3-319-46672-9
eBook Packages: Computer ScienceComputer Science (R0)