| Paper: | MLSP-L2.5 | ||
| Session: | Kernel Machines | ||
| Time: | Wednesday, May 17, 17:50 - 18:10 | ||
| Presentation: | Lecture | ||
| Topic: | Machine Learning for Signal Processing: Graphical and kernel models | ||
| Title: | Generalized Linear Kernels for One-Versus-All Classification: Application to Speaker Recognition | ||
| Authors: | Andrew Hatch, University of California, Berkeley, United States; Andreas Stolcke, SRI International, United States | ||
| Abstract: | In this paper, we examine the problem of kernel selection for one-versus-all (OVA) classification of multiclass data with support vector machines (SVMs). We focus specifically on the problem of training generalized linear kernels of the form, $k(x,y) = x^{T}Ry$, where $R$ is a positive semidefinite matrix. Our approach for training $k(x,y)$ involves first constructing a set of upper bounds on the rates of false positives and false negatives at a given score threshold. Under various conditions, minimizing these bounds leads to the closed-form solution, $R = W^{-1}$, where $W$ is the expected within-class covariance matrix of the data. We tested various parameterizations of $R$, including a diagonal parameterization that simply performs per-feature variance normalization, on the 1-conversation training condition of the SRE-2003 and SRE-2004 speaker recognition tasks. In experiments on a state-of-the-art MLLR-SVM speaker recognition system \cite{Stolcke}, the parameterization, $R = \hat{W}_{s}^{-1}$, where $\hat{W}_{s}$ is a smoothed estimate of $W$, achieves relative reductions in the minimum decision cost function (DCF) of up to 22\% below the results obtained when $R$ does per-feature variance normalization. | ||
©2018 Conference Management Services, Inc. -||- email: webmaster@icassp2006.org -||- Last updated Friday, August 17, 2012