Regularization and Sparsity for Adversarial Robustness and Stable Attribution.

Daniel Schwartz Yigit Alparslan Edward Kim

Published in: ISVC (1) (2020)

Keyphrases

mixed norm
high dimensional
multi class
regularization parameter
group lasso
learning algorithm
feature selection
data sets
computational efficiency
sparse approximation
high robustness
convex loss functions
sparsity inducing
data dependent
variable selection
prior information
image reconstruction