ABSTRACT
The Bayesian Source Separation model is different from the Bayesian Regression model in that the sources are unobserved and from the Bayesian Factor Analysis model in that there may be more or less sources than the observed dimension. Further, in the Bayesian Factor Analysis model, the variance of the unobserved factor score vectors is a priori assumed to be unity and diagonal for Psychologic reasons. The Bayesian Source Separation model allows the covariance matrix for the unobserved sources to have arbitrary variances. In the Bayesian approach to statistical inference, available prior information either from subjective expert experience, or prior experiments, is incorporated into the inferences. In the Bayesian Source Separation model available information regarding values of the model parameters is quantified using Conjugate prior distributions. Although the main focus after having performed a Bayesian Source Separation is the separated sources, there are others. The mixing coefficients are the amplitudes which determine the relative contribution of the sources.
