Abstract
Kernel principal component analysis (KPCA) is a widely-used statistical method for representation learning, where PCA is performed in reproducing kernel Hilbert space (RKHS) to extract nonlinear features from a set of training examples. Despite the success in various applications including face recognition, KPCA does not scale up well with the sample size, since, as in other kernel methods, it involves the eigen-decomposition of n ×n Gram matrix which is solved in ${\mathcal{O}}(n^3)$ time. Nystrom method is an approximation technique, where only a subset of size m≪n is exploited to approximate the eigenvectors of n ×n Gram matrix. In this paper we consider Nystrom method and its few modifications such as 'Nystrom KPCA ensemble' and 'Nystrom + randomized SVD' to improve the scalability of KPCA. We compare the performance of these methods in the task of learning face descriptors for face recognition.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.