One-shot Distibuted Algorithm for PCA with RBF Kernels
Authors:
- Fan HeFHFan He
- Kexin LvKLKexin Lv
- Jie YangJie Yang
- Xiaolin HuangXHXiaolin Huang
Preprint
Abstract
This letter proposes a one-shot algorithm for feature-distributed kernel PCA. Our algorithm is inspired by the dual relationship between sample-distributed and feature-distributed scenario. This interesting relationship makes it possible to establish distributed kernel PCA for feature-distributed cases from ideas in distributed PCA in sample-distributed scenario. In theoretical part, we analyze the approximation error for both linear and RBF kernels. The result suggests that when eigenvalues decay fast, the proposed algorithm gives high quality results with low communication cost. This result is also verified by numerical experiments, showing the effectiveness of our algorithm in practice.Keywords
All Related Versions
This publication has 14 references indexed in Scilit:
- A Distributed Framework for Dimensionality Reduction and DenoisingIEEE Transactions on Signal Processing, 2015
- A useful variant of the Davis–Kahan theorem for statisticiansBiometrika, 2014
- Distributed Principal Subspace Estimation in Wireless Sensor NetworksIEEE Journal of Selected Topics in Signal Processing, 2011
- Moving window kernel PCA for adaptive monitoring of nonlinear processesChemometrics and Intelligent Laboratory Systems, 2009
- An Improved Algorithm for Kernel Principal Component AnalysisNeural Processing Letters, 2005
- Classification of a large microarray data set: Algorithm comparison and analysis of drug signaturesGenome Research, 2005
- Nonlinear dynamic process monitoring based on dynamic kernel PCAChemical Engineering Science, 2004
- Gabor-based kernel pca with fractional power polynomial models for face recognitionIEEE Transactions on Pattern Analysis and Machine Intelligence, 2004
- Face recognition using kernel principal component analysisIEEE Signal Processing Letters, 2002
- Distributed Clustering Using Collective Principal Component AnalysisKnowledge and Information Systems, 2001