L_DMI: A Novel Information-theoretic Loss Function for Training Deep Nets Robust to Label Noise.

Yilun Xu Peng Cao Yuqing Kong Yizhou Wang

Published in: NeurIPS (2019)

Keyphrases

information theoretic
loss function
label noise
mutual information
active learning
support vector
bregman divergences
pairwise
supervised learning
training examples
reproducing kernel hilbert space
noise model
relative entropy
noise tolerant
training data
image registration
sample size
boosting algorithms