A two-stage classification procedure for near-infrared spectra based on multi-scale vertical energy wavelet thresholding and SVM-based gradient-recursive feature elimination.
H.-W. ChoSeung H. BaekEunseog YounMyong K. JeongA. TaylorPublished in: J. Oper. Res. Soc. (2009)
Keyphrases
- recursive feature elimination
- classification procedure
- multiscale
- support vector machine
- cancer diagnosis
- hyperplane
- feature selection
- support vector
- gene selection
- edge detection
- feature selection algorithms
- terms of classification accuracy
- wavelet coefficients
- hyperspectral
- wavelet transform
- rotation invariant
- support vector machine svm
- image processing
- machine learning
- natural images
- feature space
- feature vectors
- microarray data
- image segmentation
- knn
- multispectral
- image denoising
- k nearest neighbor
- image data
- classification accuracy
- remote sensing
- microarray
- pattern recognition
- subband
- kernel function
- data sets