On The Utility of Conditional Generation Based Mutual Information for Characterizing Adversarial Subspaces.
Chia-Yi HsuPei-Hsuan LuPin-Yu ChenChia-Mu YuPublished in: CoRR (2018)
Keyphrases
- mutual information
- information theoretic
- image registration
- canonical correlations
- similarity measure
- medical image registration
- information gain
- multi modal image registration
- utility function
- genetic algorithm
- generation process
- neural network
- high dimensional data
- high dimensional
- normalized mutual information
- feature selection
- statistical dependence
- mutual information maximization
- low dimensional
- data points
- kullback leibler divergence
- information theoretic measures
- conditional mutual information
- mutual subspace method