A fast on-line generalized eigendecomposition algorithm for time series segmentation

Author

Rao, Yadunandana N. ; Principe, Jose C.

Author_Institution

Dept. of Electr. & Comput. Eng., Florida Univ., Gainesville, FL, USA

fYear

2000

fDate

2000

Firstpage

266

Lastpage

271

Abstract

This paper presents a novel, fast converging on-line rule for generalized eigendecomposition (GED) and its application in time series segmentation. We adopt the concepts of deflation and power method to iteratively estimate the generalized eigencomponents. The algorithm is guaranteed to produce stable results. In the second half of the paper, we discuss the application of GED to segment time series. GED is tested for chaotic time series and speech. The simulation results are compared with the venerable Generalized Likelihood Ratio Test (GLR) as a benchmark to gauge performance

Keywords

chaos; eigenvalues and eigenfunctions; estimation theory; iterative methods; signal processing; speech processing; time series; Generalized Likelihood Ratio Test; chaotic time series; deflation; fast on-line generalized eigendecomposition algorithm; generalized eigencomponents; performance; power method; signal processing; simulation; speech; time series segmentation; Convergence; Covariance matrix; Eigenvalues and eigenfunctions; Filters; Gradient methods; Iterative algorithms; Linear discriminant analysis; Principal component analysis; Signal processing algorithms; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Adaptive Systems for Signal Processing, Communications, and Control Symposium 2000. AS-SPCC. The IEEE 2000

Conference_Location

Lake Louise, Alta.

Print_ISBN

0-7803-5800-7

Type

conf

DOI

10.1109/ASSPCC.2000.882483

Filename

882483