Kernel principal component analysis

Huang SY, Yeh YR, Eguchi S, 2009. Robust kernel principal component analysis. Neural Comput 21,3179-213. 

Scholkopf, B., Smola, A. Muller, K.R. 1999. Kernel principal components analysis. In B. Scholkopf, C. Burges A. Smola (eds.). Advances in Kernel Methods - Support Vector Learning 327-352. Cambridge (MA) MIT Press. 

HE, Q. KONG, E YAN, R. (2007) Subspace-based gearbox condition monitoring by kernel principal component analysis. Mechanical Systems and Signal Processing, 21 (4), pp. 1755-1772. 

Dimensionality reduction and data visualization Kernel Principal Component Analysis (KPCA) [6], Kernel Feature Analysis (KFA) [77] Drawing maps of chemical space... 

Mirgolbabaei, H., Echekki, T. Nonlinear reduction of combustion composition space with kernel principal component analysis. Combust. Elame 161, 118—126 (2014)... 

Gist, http //microarray.cpmc.columbia.edu/gist/. Gist is a C implementation of support vector machine classification and kernel principal components analysis. The SVM part of Gist is available as an interactive Web server at http //svm.sdsc.edu. It is a very convenient server for users who want to experiment with small datasets (hundreds of patterns). Kernels available include linear, polynomial, and radial. 

Discriminant Analysis was carried out for pure virgin coconut oil and the blended samples and a classification into two groups, pure virgin coconut oil and adulterated oils was observed. The model demonstrated the classification of pure virgin coconut oil with addition of 1-50% of palm kernel olein usinglO principal components. Therefore, detection of adulteration was possible down to 1%. 

Wet granulation and direct compression are two methods used to manufacture tablets in the pharmaceutical industry. Zomer et al. used pyrolysis-gas chromatography-mass-spectrometry coupled with SVM classification to discriminate between the two tablet production methods.Mass spectra data were submitted to a PCA analysis, and the first principal components were used as input for SVM models having linear, polynomial, and Gaussian RBF kernels. SVM classifiers with polynomial and RBF kernels performed better in prediction than discriminant analysis. 

Seven types of espresso coffee were classified by Pardo and Sberveglieri with a system composed of an electronic nose and an SVM with polynomial and Gaussian RBF kernels.For each coffee type, 36 measurements were performed with an electronic nose equipped with five thin-film semiconductor sensors based on SnOi and Ti-Fe. The output signal from sensors was submitted to a PGA analysis whose principal components (between 2 and 5) represented the input data for the SVM classifier. The error surface corresponding to various kernel parameters and number of input principal components was investigated. 

---