Variational Bayesian learning of ICA with missing data

Authors:
Kwokleung Chan;Te-Won Lee;Terrence J. Sejnowski
Affiliations:
Computational Neurobiology Laboratory, Salk Institute, La Jolla, CA;Institute for Neural Computation, University of California at San Diego, La Jolla, CA;Computational Neurobiology Laboratory, Salk Institute, La Jolla, CA and Department of Biology, University of California at San Diego, La Jolla, CA
Venue:
Neural Computation
Year:
2003

Citing 7
Cited 7

Statistical analysis with missing data

Statistical analysis with missing data
The nature of statistical learning theory

The nature of statistical learning theory
Independent factor analysis

Neural Computation
An Introduction to Variational Methods for Graphical Models

Machine Learning
Neural Networks for Pattern Recognition

Neural Networks for Pattern Recognition
Learning from Incomplete Data

Learning from Incomplete Data
Variational learning of clusters of undercomplete nonsymmetric independent components

The Journal of Machine Learning Research

On the Effect of the Form of the Posterior Approximation in Variational Learning of ICA Models

Neural Processing Letters
The Cocktail Party Problem

Neural Computation
A Gaussian mixture model for underdetermined independent component analysis

Signal Processing
Variational and stochastic inference for Bayesian source separation

Digital Signal Processing
Blind separation of nonlinear mixtures by variational Bayesian learning

Digital Signal Processing
Impact of imputation of missing values on classification error for discrete data

Pattern Recognition
Separation theorem for independent subspace analysis and its consequences

Pattern Recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

Missing data are common in real-world data sets and are a problem for many estimation techniques. We have developed a variational Bayesian method to perform independent component analysis (ICA) on high-dimensional data containing missing entries. Missing data are handled naturally in the Bayesian framework by integrating the generative density model. Modeling the distributions of the independent sources with mixture of gaussians allows sources to be estimated with different kurtosis and skewness. Unlike the maximum likelihood approach, the variational Bayesian method automatically determines the dimensionality of the data and yields an accurate density model for the observed data without overfitting problems. The technique is also extended to the clusters of ICA and supervised classification framework.