论文信息 - A Bayesian approach to blind source separation

A Bayesian approach to blind source separation

Abstract This paper presents a Bayesian statistical approach to the blind source separation problem. The blind source separation model is described; the source distribution is discussed; other approaches such as Principal Components, Independent Components, and Factor Analysis are detailed; prior distributions are introduced to incorporate available prior knowledge; the posterior distribution for the model parameters (including the number of sources) is derived; and the parameter estimation procedure is outlined. Finally Bayesian blind source separation is applied in a simulated example and its advantages over the other methods are stated.

D. Rowe

[1] D. Lawley. VI.—The Estimation of Factor Loadings by the Method of Maximum Likelihood , 1940 .

[2] Herman Rubin,et al. Statistical Inference in Factor Analysis , 1956 .

[3] Josef Guttmann,et al. Bayesian Inference in Factor Analysis -- Revised , 1973 .

[4] S. J. Press,et al. Applied multivariate analysis : using Bayesian and frequentist methods of inference , 1984 .

[5] S. J. Press,et al. Bayesian Inference in Factor Analysis , 1989 .

[6] F. Guess. Bayesian Statistics: Principles, Models, and Applications , 1990 .

[7] Terrence J. Sejnowski,et al. An Information-Maximization Approach to Blind Separation and Blind Deconvolution , 1995, Neural Computation.

[8] C. J.,et al. Maximum Likelihood and Covariant Algorithms for Independent Component Analysis , 1996 .

[9] Kevin H. Knuth,et al. Bayesian source separation and localization , 1998, Optics & Photonics.

[10] S. J. Press,et al. Robustness of bayesian factor analysis estimates , 1998 .

[11] Kevin H. Knuth. A Bayesian approach to source separation , 1999 .