Separation of uncorrelated stationary time series using autocovariance matrices

DOI10.1111/JTSA.12159zbMATH Open1381.62250arXiv1405.3388OpenAlexW1910190001WikidataQ59682307 ScholiaQ59682307MaRDI QIDQ2802912FDOQ2802912

Authors: Jari Miettinen, Katrin Illner, Klaus Nordhausen, Hannu Oja, Sara Taskinen, Fabian J. Theis 

Publication date: 3 May 2016

Published in: Journal of Time Series Analysis (Search for Journal in Brave)

Abstract: Blind source separation (BSS) is a signal processing tool, which is widely used in various fields. Examples include biomedical signal separation, brain imaging and economic time series applications. In BSS, one assumes that the observed $p$ time series are linear combinations of $p$ latent uncorrelated weakly stationary time series. The aim is then to find an estimate for an unmixing matrix, which transforms the observed time series back to uncorrelated latent time series. In SOBI (Second Order Blind Identification) joint diagonalization of the covariance matrix and autocovariance matrices with several lags is used to estimate the unmixing matrix. The rows of an unmixing matrix can be derived either one by one (deflation-based approach) or simultaneously (symmetric approach). The latter of these approaches is well-known especially in signal processing literature, however, the rigorous analysis of its statistical properties has been missing so far. In this paper, we fill this gap and investigate the statistical properties of the symmetric SOBI estimate in detail and find its limiting distribution under general conditions. The asymptotical efficiencies of symmetric SOBI estimate are compared to those of recently introduced deflation-based SOBI estimate under general multivariate MA$(infty)$ processes. The theory is illustrated by some finite-sample simulation studies as well as a real EEG data example.

Full work available at URL: https://arxiv.org/abs/1405.3388

Recommendations

• Deflation-based separation of uncorrelated stationary time series
• A more efficient second order blind identification method for separation of uncorrelated stationary time series
• Statistical properties of a blind source separation estimator for stationary time series
• On robustifying some second order blind source separation methods for nonstationary time series
• On statistical properties of blind source separation methods based on joint diagonalization

zbMATH Keywords

asymptotic normalityblind source separationjoint diagonalizationlinear processSOBI

Mathematics Subject Classification ID

Asymptotic properties of parametric estimators (62F12) Time series, auto-correlation, regression, etc. in statistics (GARCH) (62M10)

Cites Work

• Time series: theory and methods.
• Determining the Number of Factors in Approximate Factor Models
• Dynamic factor models with infinite-dimensional factor spaces: one-sided representations
• \(R\)-estimation for asymmetric independent component analysis
• Semiparametrically efficient inference based on signed ranks in symmetric independent component models
• Dynamic orthogonal components for multivariate time series
• Efficient independent component analysis
• Consistent noisy independent component analysis
• Characteristics of multivariate distributions and the invariant coordinate system
• Independent component analysis via nonparametric maximum likelihood estimation
• Independent component analysis involving autocorrelated sources with an application to functional magnetic resonance imaging
• Deflation-based separation of uncorrelated stationary time series
• Statistical properties of a blind source separation estimator for stationary time series
• Portfolio value at risk based on independent component analysis
• A stochastic algorithm for probabilistic independent component analysis
• Blind source separation for spatial compositional data
• New independent component analysis tools for time series
• On robustifying some second order blind source separation methods for nonstationary time series

Cited In (18)

• Fourth moments and independent component analysis
• Sampling properties of color independent component analysis
• Time Series Source Separation Using Dynamic Mode Decomposition
• A more efficient second order blind identification method for separation of uncorrelated stationary time series
• Stationary subspace analysis based on second-order statistics
• Deflation-based separation of uncorrelated stationary time series
• Sliced average variance estimation for multivariate time series
• Statistical properties of a blind source separation estimator for stationary time series
• Blind source separation for spatial compositional data
• Extracting conditionally heteroskedastic components using independent component analysis
• New independent component analysis tools for time series
• Determining the signal dimension in second order source separation
• On the usage of joint diagonalization in multivariate statistics
• Blind source separation for compositional time series
• Modeling temporally uncorrelated components of complex-valued stationary processes
• On robustifying some second order blind source separation methods for nonstationary time series
• Decomposition of stationary time series - finding the highly correlated components of stochastic processes
• Independent component analysis: a statistical perspective

Uses Software

• BSSasymp

This page was built for publication: Separation of uncorrelated stationary time series using autocovariance matrices

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2802912)