TY - JOUR
T1 - A new discriminative Kernel from probabilistic models
AU - Tsuda, Koji
AU - Kawanabe, Motoaki
AU - Rätsch, Gunnar
AU - Sonnenburg, Sören
AU - Müller, Klaus Robert
N1 - Copyright:
Copyright 2005 Elsevier B.V., All rights reserved.
PY - 2002/10
Y1 - 2002/10
N2 - Recently, Jaakkola and Haussler (1999) proposed a method for constructing kernel functions from probabilistic models. Their so-called Fisher kernel has been combined with discriminative classifiers such as support vector machines and applied successfully in, for example, DNA and protein analysis. Whereas the Fisher kernel is calculated from the marginal log-likelihood, we propose the TOP kernel derived from tangent vectors of posterior log-odds. Furthermore, we develop a theoretical framework on feature extractors from probabilistic models and use it for analyzing the TOP kernel. In experiments, our new discriminative TOP kernel compares favorably to the Fisher kernel.
AB - Recently, Jaakkola and Haussler (1999) proposed a method for constructing kernel functions from probabilistic models. Their so-called Fisher kernel has been combined with discriminative classifiers such as support vector machines and applied successfully in, for example, DNA and protein analysis. Whereas the Fisher kernel is calculated from the marginal log-likelihood, we propose the TOP kernel derived from tangent vectors of posterior log-odds. Furthermore, we develop a theoretical framework on feature extractors from probabilistic models and use it for analyzing the TOP kernel. In experiments, our new discriminative TOP kernel compares favorably to the Fisher kernel.
UR - http://www.scopus.com/inward/record.url?scp=0036780246&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0036780246&partnerID=8YFLogxK
U2 - 10.1162/08997660260293274
DO - 10.1162/08997660260293274
M3 - Article
C2 - 12396568
AN - SCOPUS:0036780246
VL - 14
SP - 2397
EP - 2414
JO - Neural Computation
JF - Neural Computation
SN - 0899-7667
IS - 10
ER -