Information-theoretic atomic representation for robust pattern classification.
Yulong WangYuan Yan TangLuoqing LiPatrick Shen-Pei WangPublished in: ICPR (2016)
Keyphrases
- information theoretic
- pattern classification
- information theory
- mutual information
- pattern recognition
- information theoretic measures
- theoretic framework
- information bottleneck
- kullback leibler divergence
- feature extraction
- jensen shannon divergence
- nearest neighbor rule
- pattern classification problems
- relative entropy
- log likelihood
- entropy measure
- kl divergence
- image representation
- pairwise
- computer vision
- maximum likelihood
- image classification
- probabilistic model
- distributional clustering
- jensen shannon