Information-Theoretic Feature Selection for Human Micro-Doppler Signature Classification.
Burkan TekeliSevgi Zubeyde GurbuzMelda YukselPublished in: IEEE Trans. Geosci. Remote. Sens. (2016)
Keyphrases
- information theoretic
- feature selection
- mutual information
- information theory
- classification accuracy
- jensen shannon divergence
- theoretic framework
- support vector
- text classification
- multi modality
- entropy measure
- information theoretic measures
- feature extraction
- feature selection algorithms
- kullback leibler divergence
- support vector machine
- information bottleneck
- feature set
- feature vectors
- feature space
- machine learning
- decision trees
- log likelihood
- relative entropy
- distributional clustering
- minimum description length
- text categorization
- pattern recognition
- kl divergence
- selection criterion
- bregman divergences
- naive bayes
- multi class
- knn
- training set