Bayesian source separation of functional sources
Daniel B. Rowe · Journal of Interdisciplinary Mathematics · 2003
This paper incorporates available prior knowledge of the source waveforms into the Bayesian approach to blind source separation. The source separation model is described, prior distributions are introduced to quantify available prior knowledge regarding the model parameters, the posterior distribution for the model parameters is formed, and parameter estimation is detailed. Finally, the methods discussed are applied to an example.