Login / Signup
Training a network to attend like human drivers saves it from common but misleading loss functions.
Ye Xia
Danqing Zhang
Alexei Pozdnoukhov
Ken Nakayama
Karl Zipser
David Whitney
Published in:
CoRR (2017)
Keyphrases
</>
loss function
support vector
pairwise
squared error
logistic regression
learning to rank
loss minimization
training set
reproducing kernel hilbert space
bregman divergences
hinge loss
partially labeled data
special case
training samples
risk minimization
update rules