Login / Signup
Tackling Early Sparse Gradients in Softmax Activation Using Leaky Squared Euclidean Distance.
Wei Shen
Rujie Liu
Published in:
CoRR (2018)
Keyphrases
</>
squared euclidean distance
relative entropy
high dimensional
k means
sparse representation
information theoretic
bregman divergences
clustering algorithm
parameter estimation
maximum entropy