Information-theoretic criteria for the design of compressive subspace classifiers.
Matthew S. NoklebyMiguel R. D. RodriguesA. Robert CalderbankPublished in: ICASSP (2014)
Keyphrases
- probabilistic model
- information theoretic
- mutual information
- information theory
- theoretic framework
- information theoretic measures
- multi modality
- jensen shannon divergence
- feature selection
- compressive sensing
- information bottleneck
- high dimensional
- training data
- decision trees
- relative entropy
- image analysis
- maximum likelihood
- log likelihood
- kullback leibler divergence