The Leave-One-Out Kernel
Recently, several attempts have been made for deriving data-dependent kernels from distribution estimates with parametric models (e.g. the Fisher kernel). In this paper, we propose a new kernel derived from any distribution estimators, parametric or nonparametric. This kernel is called the Leave-one-out kernel (i.e. LOO kernel), because the leave-one-out process plays an important role to compute this kernel. We will show that, when applied to a parametric model, the LOO kernel converges to the Fisher kernel asymptotically as the number of samples goes to infinity.
Unable to display preview. Download preview PDF.
- 1.S. Amari and H. Nagaoka. Methods of Information Geometry, volume 191 of Translations of Mathematical Monographs. American Mathematical Society, 2001.Google Scholar
- 3.T.S. Jaakkola, M. Diekhans, and D. Haussler. A discriminative framework for detecting remote protein homologies. J. Biol., 7:95–114, 2000.Google Scholar
- 4.T.S. Jaakkola and D. Haussler. Exploiting generative models in discriminative classifiers. In NIPS 11, pages 487–493. MIT Press, 1999.Google Scholar
- 5.A.K. Jain and R.C. Dubes. Algorithms for Clustering Data. Prentice Hall, 1988.Google Scholar
- 7.M. Rattray. A model-based distance for clustering. In Proc. IJCNN’00, 2000.Google Scholar
- 8.M.E. Tipping. Deriving cluster analytic distance functions from gaussian mixture models. In D. Willshaw and A. Murray, editors, Proceedings of ICANN’99, pages 815–820. IEE Press, 1999.Google Scholar