Recently, several attempts have been made for deriving datadependent kernels from distribution estimates withparametric 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. © Springer-Verlag Berlin Heidelberg 2002.
CITATION STYLE
Tsuda, K., & Kawanabe, M. (2002). The leave-one-out kernel. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2415 LNCS, pp. 727–732). Springer Verlag. https://doi.org/10.1007/3-540-46084-5_118
Mendeley helps you to discover research relevant for your work.