Nonlinear Component Analysis as a Kernel Eigenvalue Problem

@techreport{1509,
  title = {Nonlinear Component Analysis as a Kernel Eigenvalue Problem},
  abstract = {We describe a new method for performing a nonlinear form of Principal Component Analysis. By the use of integral operator kernel functions, we can efficiently compute principal components in high-dimensional feature spaces,
  related to input space by some nonlinear map; for instance the space of all possible 5-pixel products in 16 x 16 images.  We give the derivation of the method, along with a discussion of other techniques which can be made nonlinear with the kernel approach; and present first experimental results on nonlinear feature extraction for pattern recognition.},
  number = {44},
  organization = {Max-Planck-Gesellschaft},
  institution = {Max Planck Institute for Biological Cybernetics Tübingen},
  school = {Biologische Kybernetik},
  month = dec,
  year = {1996},
  note = {This technical report has also been <a href="/main/publication.php?machwas=view_e&edit_lfnr=730">published elsewhere</a>},
  author = {Sch{\"o}lkopf, B. and Smola, AJ. and M{\"u}ller, K-R.},
  month_numeric = {12}
}