Variational Bayes for non-Gaussian autoregressive models

We describe a variational Bayesian (VB) learning algorithm for Non-Gaussian Autoregressive (AR) models. The noise is modelled as a mixture of Gaussians rather than the usual single Gaussian. This allows different data points to be associated with different noise levels and effectively provides a robust estimation of AR coefficients. The VB framework is used to prevent overfitting and provides model order selection criteria both for AR order and noise model order. The algorithm is applied to synthetic data and to EEG.

[1]  Daphne N. Yu,et al.  High-resolution EEG mapping of cortical activation related to working memory: effects of task difficulty, type of processing, and practice. , 1997, Cerebral cortex.

[2]  Hagai Attias,et al.  A Variational Bayesian Framework for Graphical Models , 1999 .

[3]  Stephen J. Roberts,et al.  An ensemble learning approach to independent component analysis , 2000, Neural Networks for Signal Processing X. Proceedings of the 2000 IEEE Signal Processing Society Workshop (Cat. No.00TH8501).

[4]  S. Roberts,et al.  Bayesian methods for autoregressive models , 2000, Neural Networks for Signal Processing X. Proceedings of the 2000 IEEE Signal Processing Society Workshop (Cat. No.00TH8501).