Q-MKL: Matrix-induced Regularization in Multi-Kernel Learning with Applications to Neuroimaging.

Chris Hinrichs Vikas Singh Jiming Peng Sterling C. Johnson

Published in: NIPS (2012)

Keyphrases

kernel learning
kernel matrices
multiple kernel learning
kernel matrix
kernel based learning methods
kernel function
kernel methods

optimal kernel
gaussian kernels
support vector regression
semi supervised
low rank
semidefinite program
multiple kernel

linear combination
pairwise constraints
base kernels
feature selection
reproducing kernel hilbert space
input space
learning tasks

support vector
binary classification
metric learning
multi task
multi class
machine learning