Nonlinear Component Analysis as a Kernel Eigenvalue Problem
- 1 July 1998
- journal article
- Published by MIT Press in Neural Computation
- Vol. 10 (5) , 1299-1319
- https://doi.org/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 × 16 images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.Keywords
This publication has 6 references indexed in Scilit:
- Support-vector networksMachine Learning, 1995
- Application of the Karhunen-Loeve procedure for the characterization of human facesPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1990
- Backpropagation Applied to Handwritten Zip Code RecognitionNeural Computation, 1989
- Principal CurvesJournal of the American Statistical Association, 1989
- Simplified neuron model as a principal component analyzerJournal of Mathematical Biology, 1982
- On optimal nonlinear associative recallBiological Cybernetics, 1975