Tensor-based total bregman divergences between graphs.

Francisco Escolano Meizhu Liu Edwin R. Hancock

Published in: ICCV Workshops (2011)

Keyphrases

bregman divergences
cost sensitive
maximum entropy
information theoretic
mahalanobis distance
exponential family
nearest neighbor
special case
loss function
learning theory
theoretical guarantees
kl divergence
feature space
active learning