Modifications of most expressive feature reordering criteria for supervised kernel Principal Component Analysis.
Krzysztof AdamiakPiotr DuchDominik ZurekKrzysztof SlotPublished in: CYBCONF (2015)
Keyphrases
- high dimensional
- kernel principal component analysis
- kernel function
- feature space
- input space
- feature extraction
- pattern recognition
- high dimensional data
- feature vectors
- semi supervised
- kernel methods
- unsupervised learning
- feature set
- learning algorithm
- machine learning
- supervised learning
- principal component analysis
- data sets
- image features
- pairwise
- preprocessing
- similarity measure
- discriminant analysis
- gabor filters
- kernel matrix