On Using Prototype Reduction Schemes and Classifier Fusion Strategies to Optimize Kernel-Based Nonlinear Subspace Methods.
Sang-Woon KimB. John OommenPublished in: IEEE Trans. Pattern Anal. Mach. Intell. (2005)
Keyphrases
- fusion strategies
- kernel based nonlinear
- subspace methods
- fusion methods
- fusion method
- hyperspectral imagery
- feature space
- kernel function
- face recognition
- linear discriminant analysis
- ensemble methods
- principal component analysis
- kernel trick
- support vector machine
- data fusion
- lower dimensional
- dimensionality reduction
- combining multiple
- blind source separation
- support vector
- video sequences
- training data
- feature extraction