Abstract
For many biomedical modelling tasks a number of different types of data may influence predictions made by the model. An established approach to pursuing supervised learning with multiple types of data is to encode these different types of data into separate kernels and use multiple kernel learning. In this paper we propose a simple iterative approach to multiple kernel learning (MKL), focusing on multi-class classification. This approach uses a block L 1-regularization term leading to a jointly convex formulation. It solves a standard multi-class classification problem for a single kernel, and then updates the kernel combinatorial coefficients based on mixed RKHS norms. As opposed to other MKL approaches, our iterative approach delivers a largely ignored message that MKL does not require sophisticated optimization methods while keeping competitive training times and accuracy across a variety of problems. We show that the proposed method outperforms state-of-the-art results on an important protein fold prediction dataset and gives competitive performance on a protein subcellular localization task.
Chapter PDF
Similar content being viewed by others
Keywords
References
Argyriou, A., Evgeniou, T., Pontil, M.: Multi-task feature learning. In: NIPS (2006)
Argyriou, A., Micchelli, C.A., Pontil, M., Ying, Y.: A spectral regularization framework for multi-task structure learning. In: NIPS (2007)
Bach, F., Lanckriet, G.R.G., Jordan, M.I.: Multiple kernel learning, conic duality and the SMO algorithm. In: ICML (2004)
Boyd, S., Vandenberghe, L.: Convex Optimization. Cambridge University Press, Cambridge (2004)
Damoulas, T., Girolami, M.: Probabilistic multi-class multi-kernel learning: On protein fold recognition and remote homology detection. Bioinformatics 24(10), 1264–1270 (2008)
Ding, C., Dubchak, I.: Multi-class protein fold recognition using support vector machines and neural networks. Bioinformatics 17, 349–358 (2001)
Gardy, J.L., et al.: PSORTb v.2.0: expanded prediction of bacterial protein subcellular localization and insights gained from comparative proteome analysis. Bioinformatics 21, 617–623 (2004)
Girolami, M., Rogers, S.: Hierarchic Bayesian models for kernel learning. In: ICML (2005)
Lanckriet, G.R.G., Cristianini, N., Bartlett, P., Ghaoui, L.E., Jordan, M.I.: Learning the kernel matrix with semidefinite programming. J. of Machine Learning Research 5, 27–72 (2004)
Lin, Y., Zhang, H.: Component selection and smoothing in multivariate nonparametric regression. Annals of Statistics 34, 2272–2297 (2006)
Micchelli, C.A., Pontil, M.: Learning the kernel function via regularization. J. of Machine Learning Research 6, 1099–1125 (2005)
Ong, C.S., Smola, A.J., Williamson, R.C.: Learning the kernel with hyperkernels. J. of Machine Learning Research 6, 1043–1071 (2005)
Rakotomamonjy, A., Bach, F., Canu, S., Grandvalet, Y.: More efficiency in multiple kernel learning. In: ICML (2007)
Rakotomamonjy, A., Bach, F., Canu, S., Grandvalet, Y.: SimpleMKL. J. of Machine Learning Research 9, 2491–2521 (2008)
Schölkopf, B., Smola, A.J.: Learning with Kernels. MIT Press, Cambridge (2002)
Shawe-Taylor, J., Cristianini, N.: Kernel methods for pattern analysis. Cambridge University Press, Cambridge (2004)
Shen, H.B., Chou, K.C.: Ensemble classifier for protein fold pattern recognition. Bioinformatics 22, 1717–1722 (2006)
Sonnenburg, S., Rätsch, G., Schäfer, C., Schölkopf, B.: Large scale multiple kernel learning. J. of Machine Learning Research 7, 1531–1565 (2006)
Ye, J., Ji, S., Chen, J.: Multi-class discriminant kernel learning via convex programming. J. of Machine Learning Research 9, 719–758 (2008)
Zien, A., Ong, C.: Multi-class multiple kernel learning. In: ICML (2007)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Ying, Y., Campbell, C., Damoulas, T., Girolami, M. (2009). Class Prediction from Disparate Biological Data Sources Using an Iterative Multi-Kernel Algorithm. In: Kadirkamanathan, V., Sanguinetti, G., Girolami, M., Niranjan, M., Noirel, J. (eds) Pattern Recognition in Bioinformatics. PRIB 2009. Lecture Notes in Computer Science(), vol 5780. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-04031-3_37
Download citation
DOI: https://doi.org/10.1007/978-3-642-04031-3_37
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-04030-6
Online ISBN: 978-3-642-04031-3
eBook Packages: Computer ScienceComputer Science (R0)