123
click to vote
CORR
2008
Springer
15 years 2 months ago
2008
Springer
In this paper, we propose a method for support vector machine classification using indefinite kernels. Instead of directly minimizing or stabilizing a nonconvex loss function, our...