Information-Theoretic Bounds on The Removal of Attribute-Specific Bias From Neural Networks.
Jiazhi LiMahyar KhayatkhoeiJiageng ZhuHanchen XieMohamed E. HusseinWael AbdAlmageedPublished in: CoRR (2023)
Keyphrases
- information theoretic
- neural network
- information theory
- mutual information
- theoretic framework
- competitive learning
- information bottleneck
- information theoretic measures
- log likelihood
- multi modality
- relative entropy
- jensen shannon divergence
- entropy measure
- kullback leibler divergence
- upper bound
- pattern recognition
- attribute values
- worst case
- minimum description length
- distributional clustering
- image segmentation