Nonlinear Component Analysis as a Kernel Eigenvalue Problem. Schölkopf, B., Smola, A., & Müller, K. Neural Computation, 10(5):1299–1319, July, 1998. doi abstract bibtex A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one 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 five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition. A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one 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 five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.
@article{scholkopfNonlinearComponentAnalysis1998,
title = {Nonlinear {{Component Analysis}} as a {{Kernel Eigenvalue Problem}}},
author = {Sch{\"o}lkopf, Bernhard and Smola, Alexander and M{\"u}ller, Klaus-Robert},
year = {1998},
month = jul,
volume = {10},
pages = {1299--1319},
issn = {0899-7667},
doi = {10.1162/089976698300017467},
abstract = {A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one 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 five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition. A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one 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 five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.},
journal = {Neural Computation},
keywords = {*imported-from-citeulike-INRMM,~INRMM-MiD:c-3474398,high-impact-publication,mathematics,non-linearity,nonlinear-correlation,pca},
lccn = {INRMM-MiD:c-3474398},
number = {5}
}
Downloads: 0
{"_id":"6GapW8wyDA6kEF5G2","bibbaseid":"schlkopf-smola-mller-nonlinearcomponentanalysisasakerneleigenvalueproblem-1998","downloads":0,"creationDate":"2016-06-22T10:19:51.440Z","title":"Nonlinear Component Analysis as a Kernel Eigenvalue Problem","author_short":["Schölkopf, B.","Smola, A.","Müller, K."],"year":1998,"bibtype":"article","biburl":"https://sharefast.me/php/download.php?id=zOUKvA&token=29","bibdata":{"bibtype":"article","type":"article","title":"Nonlinear Component Analysis as a Kernel Eigenvalue Problem","author":[{"propositions":[],"lastnames":["Schölkopf"],"firstnames":["Bernhard"],"suffixes":[]},{"propositions":[],"lastnames":["Smola"],"firstnames":["Alexander"],"suffixes":[]},{"propositions":[],"lastnames":["Müller"],"firstnames":["Klaus-Robert"],"suffixes":[]}],"year":"1998","month":"July","volume":"10","pages":"1299–1319","issn":"0899-7667","doi":"10.1162/089976698300017467","abstract":"A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one 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 five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition. A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one 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 five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.","journal":"Neural Computation","keywords":"*imported-from-citeulike-INRMM,~INRMM-MiD:c-3474398,high-impact-publication,mathematics,non-linearity,nonlinear-correlation,pca","lccn":"INRMM-MiD:c-3474398","number":"5","bibtex":"@article{scholkopfNonlinearComponentAnalysis1998,\n title = {Nonlinear {{Component Analysis}} as a {{Kernel Eigenvalue Problem}}},\n author = {Sch{\\\"o}lkopf, Bernhard and Smola, Alexander and M{\\\"u}ller, Klaus-Robert},\n year = {1998},\n month = jul,\n volume = {10},\n pages = {1299--1319},\n issn = {0899-7667},\n doi = {10.1162/089976698300017467},\n abstract = {A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one 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 five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition. A new method for performing a nonlinear form of principal component analysis is proposed. By the use of integral operator kernel functions, one 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 five-pixel products in 16 x 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.},\n journal = {Neural Computation},\n keywords = {*imported-from-citeulike-INRMM,~INRMM-MiD:c-3474398,high-impact-publication,mathematics,non-linearity,nonlinear-correlation,pca},\n lccn = {INRMM-MiD:c-3474398},\n number = {5}\n}\n\n","author_short":["Schölkopf, B.","Smola, A.","Müller, K."],"key":"scholkopfNonlinearComponentAnalysis1998","id":"scholkopfNonlinearComponentAnalysis1998","bibbaseid":"schlkopf-smola-mller-nonlinearcomponentanalysisasakerneleigenvalueproblem-1998","role":"author","urls":{},"keyword":["*imported-from-citeulike-INRMM","~INRMM-MiD:c-3474398","high-impact-publication","mathematics","non-linearity","nonlinear-correlation","pca"],"downloads":0},"search_terms":["nonlinear","component","analysis","kernel","eigenvalue","problem","schölkopf","smola","müller"],"keywords":["high-impact-publication","mathematics","non-linearity","nonlinear-correlation","pca","*imported-from-citeulike-inrmm","~inrmm-mid:c-3474398"],"authorIDs":[],"dataSources":["5S2zj2hKW8TWTkuMq"]}