Mutual Information Learned Classifiers: an Information-theoretic Viewpoint of Training Deep Learning Classification Systems.
Jirong YiQiaosheng ZhangZhen ChenQiao LiuWei ShaoPublished in: CoRR (2022)
Keyphrases
- classification systems
- information theoretic
- mutual information
- deep learning
- viewpoint
- deep architectures
- information theory
- information theoretic measures
- feature selection
- unsupervised learning
- image registration
- similarity measure
- jensen shannon divergence
- machine learning
- training set
- mental models
- multi modality
- kullback leibler divergence
- training samples
- weakly supervised
- higher order
- d objects
- multiple views
- training examples
- supervised learning
- object class
- training data