Fourth Moments and Independent Component Analysis

Abstract

In independent component analysis it is assumed that the components of the observed random vector are linear combinations of latent independent random variables, and the aim is then to find an estimate for a transformation matrix back to these independent components. In the engineering literature, there are several traditional estimation procedures based on the use of fourth moments, such as FOBI (fourth order blind identification), JADE (joint approximate diagonalization of eigenmatrices), and FastICA, but the statistical properties of these estimates are not well known. In this paper various independent component functionals based on the fourth moments are discussed in detail, starting with the corresponding optimization problems, deriving the estimating equations and estimation algorithms, and finding asymptotic statistical properties of the estimates. Comparisons of the asymptotic variances of the estimates in wide independent component models show that in most cases JADE and the symmetric version of FastICA perform better than their competitors.

Published at http://dx.doi.org/10.1214/15-STS520 in the Statistical Science (http://www.imstat.org/sts/) by the Institute of Mathematical Statistics (http://www.imstat.org)

References in corpus (3)

• Efficient independent component analysis
• Separation of uncorrelated stationary time series using autocovariance matrices
• Semiparametrically efficient inference based on signed ranks in symmetric independent component models

Cited by in corpus (21)

• Independent component analysis for tensor-valued data
• Nonlinear PCA for Spatio-Temporal Analysis of Earth Observation Data
• A convergence and asymptotic analysis of the generalized symmetric FastICA algorithm
• The squared symmetric FastICA estimator
• Independent component analysis for multivariate functional data
• Extracting conditionally heteroscedastic components using ICA
• Blind source separation for non-stationary random fields
• On the number of signals in multivariate time series
• JADE for Tensor-Valued Observations
• Sliced Average Variance Estimation for Multivariate Time Series
• Clustering Via Finite Nonparametric ICA Mixture Models
• Large-Sample Properties of Blind Estimation of the Linear Discriminant Using Projection Pursuit
• Numerical considerations and a new implementation for ICS
• Non-Gaussian component analysis: testing the dimension of the signal subspace
• Stationary subspace analysis based on second-order statistics
• Independent Component Analysis for Compositional Data
• Fast tensorial JADE
• Large-Sample Properties of Non-Stationary Source Separation for Gaussian Signals
• Unsupervised linear discrimination using skewness
• A method for sparse and robust independent component analysis
• Independent vector analysis -- an introduction for statisticians