Online Dominant Generalized Eigenvectors Extraction Via A Randomized Method

The generalized Hermitian eigendecomposition problem is ubiquitous in signal and machine learning applications. Considering the need of processing streaming data in practice and restrictions of existing methods, this paper is concerned with fast and efficient generalized eigenvectors tracking. We first present a computationally efficient algorithm based on randomization termed alternate-projections randomized eigenvalue decomposition (APR-EVD) to solve a standard eigenvalue problem. By exploiting rank-1 strategy, two online algorithms based on APR-EVD are developed for the dominant generalized eigenvectors extraction. Numerical examples show the practical applicability and efficacy of the proposed online algorithms.

[1]  D. R. Dennis R. Morgan,et al.  Adaptive algorithms for solving generalized eigenvalue signal enhancement problems , 2004, Signal Process..

[2]  Zhansheng Duan,et al.  Adaptive Generalized Eigen-Pairs Extraction Algorithms and Their Convergence Analysis , 2016, IEEE Transactions on Signal Processing.

[3]  Jinho Choi,et al.  A novel adaptive beamforming algorithm for antenna array CDMA systems with strong interferers , 2002, IEEE Trans. Veh. Technol..

[4]  Lucas C. Parra,et al.  Blind Source Separation via Generalized Eigenvalue Decomposition , 2003, J. Mach. Learn. Res..

[5]  D. Sorensen.  Numerical methods for large eigenvalue problems , 2002, Acta Numerica.

[6]  Haifeng Chen,et al.  Failure Detection in Large-Scale Internet Services by Principal Subspace Mapping , 2007, IEEE Transactions on Knowledge and Data Engineering.

[7]  Ming Gu,et al.  Subspace Iteration Randomization and Singular Value Problems , 2014, SIAM J. Sci. Comput..

[8]  Jacob Benesty,et al.  Signal Enhancement with Variable Span Linear Filters , 2016 .

[9]  Shifei Ding,et al.  An overview on nonparallel hyperplane support vector machine algorithms , 2013, Neural Computing and Applications.

[10]  Toshihisa Tanaka,et al.  Fast Generalized Eigenvector Tracking Based on the Power Method , 2009, IEEE Signal Processing Letters.

[11]  Isao Yamada,et al.  Adaptive Normalized Quasi-Newton Algorithms for Extraction of Generalized Eigen-Pairs and Their Convergence Analysis , 2013, IEEE Transactions on Signal Processing.

[12]  G. Stewart.  A generalization of Saad's theorem on Rayleigh–Ritz approximations , 2001 .

[13]  Ralf Möller,et al.  Coupled principal component analysis , 2004, IEEE Transactions on Neural Networks.

[14]  Y. Saad,et al.  Numerical Methods for Large Eigenvalue Problems , 2011 .

[15]  Amir Valizadeh,et al.  A constrained optimization approach for an adaptive generalized subspace tracking algorithm , 2010, Comput. Electr. Eng..

[16]  Xixuan Han,et al.  Regularized generalized eigen-decomposition with applications to sparse supervised feature extraction and sparse discriminant analysis , 2016, Pattern Recognit..

[17]  Nathan Halko,et al.  Finding Structure with Randomness: Probabilistic Algorithms for Constructing Approximate Matrix Decompositions , 2009, SIAM Rev..

[18]  Jack Dongarra,et al.  Templates for the Solution of Algebraic Eigenvalue Problems , 2000, Software, environments, tools.

[19]  Bin Yang,et al.  Projection approximation subspace tracking , 1995, IEEE Trans. Signal Process..

[20]  Samir Attallah,et al.  A Fast Adaptive Algorithm for the Generalized Symmetric Eigenvalue Problem , 2008, IEEE Signal Processing Letters.

[21]  Jian Yang,et al.  RLS-based adaptive algorithms for generalized eigen-decomposition , 2006, IEEE Transactions on Signal Processing.

[22]  Li Shen,et al.  A Decomposition Algorithm for the Sparse Generalized Eigenvalue Problem , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[23]  G. W. Stewart,et al.  Matrix Algorithms: Volume 1, Basic Decompositions , 1998 .

[24]  Jie Chen,et al.  Randomized Truncated Pivoted QLP Factorization for Low-Rank Matrix Recovery , 2019, IEEE Signal Processing Letters.

[25]  Isao Yamada,et al.  An adaptive extraction of generalized eigensubspace by using exact nested orthogonal complement structure , 2013, Multidimens. Syst. Signal Process..