Handbook of Blind Source Separation : Independent Component Analysis and Applications
- Comon, Pierre
- Burlington : Elsevier Science, 2010.
- Physical Description:
- 1 online resource (856 pages)
- Additional Creators:
- Jutten, Christian, 1954-
- Language Note:
- Front cover; Half page; Title page; Copyright page; Contents; About the editors; Preface; Contributors; Chapter 1. Introduction; 1.1. Genesis of blind source separation; 1.2. Problem formalization; 1.3. Source separation methods; 1.4. Spatial whitening, noise reduction and PCA; 1.5. Applications; 1.6. Content of the handbook; References; Chapter 2. Information; 2.1. Introduction; 2.2. Methods based on mutual information; 2.3. Methods based on mutual information rate; 2.4. Conclusion and perspectives; References; Chapter 3. Contrasts; 3.1. Introduction; 3.2. Cumulants; 3.3. MISO contrasts., 3.4. MIMO contrasts for static mixtures3.5. MIMO contrasts for dynamic mixtures; 3.6. Constructing other contrast criteria; 3.7. Conclusion; References; Chapter 4. Likelihood; 4.1. Introduction: Models and likelihood; 4.2. Transformation model and equivariance; 4.3. Independence; 4.4. Identifiability, stability, performance; 4.5. Non-Gaussian models; 4.6. Gaussian models; 4.7. Noisy models; 4.8. Conclusion: A general view; 4.9. Appendix: Proofs; References; Chapter 5. Algebraic methods after prewhitening; 5.1. Introduction; 5.2. Independent component analysis., 5.3. Diagonalization in least squares sense5.4. Simultaneous diagonalization of matrix slices; 5.5. Simultaneous diagonalization of third-order tensor slices; 5.6. Maximization of the tensor trace; References; Chapter 6. Iterative algorithms; 6.1. Introduction; 6.2. Model and goal; 6.3. Contrast functions for iterative BSS/ICA; 6.4. Iterative search algorithms: Generalities; 6.5. Iterative whitening; 6.6. Classical adaptive algorithms; 6.7. Relative (natural) gradient techniques; 6.8. Adapting the nonlinearities; 6.9. Iterative algorithms based on deflation; 6.10. The FastICA algorithm., 6.11. Iterative algorithms with optimal step size6.12. Summary, conclusions and outlook; References; Chapter 7. Second-order methods based on color; 7.1. Introduction; 7.2. WSS processes; 7.3. Problem formulation, identifiability and bounds; 7.4. Separation based on joint diagonalization; 7.5. Separation based on maximum likelihood; 7.6. Additional issues; References; Chapter 8. Convolutive mixtures; 8.1. Introduction and mixture model; 8.2. Invertibility of convolutive MIMO mixtures; 8.3. Assumptions; 8.4. Joint separating methods; 8.5. Iterative and deflation methods., and 8.6. Non-stationary contextReferences; Chapter 9. Algebraic identification of under-determined mixtures; 9.1. Observation model; 9.2. Intrinsic identifiability; 9.3. Problem formulation; 9.4. Higher-order tensors; 9.5. Tensor-based algorithms; 9.6. Appendix: expressions of complex cumulants; References; Chapter 10. Sparse component analysis; 10.1. Introduction; 10.2. Sparse signal representations; 10.3. Joint sparse representation of mixtures; 10.4. Estimating the mixing matrix by clustering; 10.5. Square mixing matrix: Relative Newton method; 10.6. Separation with a known mixing matrix.
- A key task of engineers is to design and analyse systems; however, they often have to do this without knowing a system's parameters. BSS is a very important area in signal processing as it enables engineers to derive the unknown inputs of a system from its known outputs. It also enables the separation of a set of signals from mixed set of signals. This is particularly important in telecommunications and biomedical engineering but is also key in speech, acoustic, audio and music processing and scientific data analysis. It is, therefore, a method that has wide applicability and is a very useful.
- 10.7. Conclusion.
View MARC record | catkey: 37443785